The rapid proliferation of large audio models (LAMs) demands efficient approaches for model comparison, yet comprehensive benchmarks are costly.To fill this gap, we investigate whether minimal subsets can reliably evaluate LAMs while reducing costs and data redundancy.Analyzing 10 subset selection methods with 18 audio models across 40 tasks covering major LAM evaluation dimensions, we show that subsets of just 50 examples (0.3% of data) can achieve over 0.93 Pearson correlation with full benchmark scores.To understand how well these scores align with what practitioners ultimately care about-user satisfaction-we collect 776 human preference ratings from realistic voice assistant conversations, finding that both subsets and full benchmark achieve only 0.85 correlation with human.To better predict preferences, we trained regression models on these selected subsets, achieving 0.98 correlation-outperforming regression models trained on both random subsets and the full benchmark.This demonstrates that in regression modeling, well-curated subsets outpredict the full benchmark, showing quality over quantity.We opensource these regression-weighted subsets as the HUMANS benchmark, an efficient proxy for LAM evaluation that captures both benchmark performance and user preferences.

Large language models promise a broad set of functions, but when not given a specific objective, they default to generic results. We demonstrate that inferring the user’s in-the-moment objective, then rapidly optimizing for that singular objective, enables LLMs to produce specialized tools, interfaces, and responses. Our work introduces just-in-time objectives, which model a user’s goals to specialize LLM systems on the fly. We contribute an architecture for automatically inducing such objectives by passively observing user behavior, then steering downstream AI systems through generation and evaluation against this objective. Inducing just-in-time objectives (e.g., “Clarify the abstract’s research contribution”) enables automatic generation of tools, e.g., those that critique a draft based on relevant HCI methodologies, anticipate related researchers’ reactions, or surface ambiguous terminology. In a series of experiments on participants’ own tasks, JIT objectives enable LLM outputs that achieve 66–86% win rates over typical LLMs. In-person use sessions confirm that JIT objectives produce specialized tools that are unique to each participant and are rated as significantly higher quality than a standard LLM chat tool.

The growing demand for accessible mental health support requires training more counselors, yet existing approaches remain resource-intensive and difficult to scale. LLMs can realistically simulate patients and generate actionable feedback for training, but their actual impact on novice counselor skill development remains unknown. We developed an LLM-simulated practice and feedback system and conducted a randomized study with 94 novice counselors, comparing practice alone versus practice with feedback. We evaluated behavioral performance, self-efficacy, and qualitative reflections. Results showed the practice-and-feedback group improved in client-centered microskills (reflections, questions), while the practice-alone group showed no improvements. For empathy, the practice-alone group declined over time and performed significantly worse than the feedback group. Qualitative interviews reinforced these findings: feedback helped participants adopt a client-centered listening approach, while practice-alone participants remained solution-oriented. These results suggest LLM-based training systems can promote effective skill development, and combining simulated practice with structured feedback is critical for meaningful improvement.

Although generative AI is being deployed into classrooms with promises of aiding teachers, educators caution that these tools can have unintended pedagogical repercussions, including cultural misrepresentation and bias. These concerns are heightened in low-resource language and Indigenous education settings, where AI systems frequently underperform. We investigate these challenges in Hawai‘i, where public schools operate under a statewide mandate to integrate Hawaiian language and culture into education. Through four co-design workshops with 22 public school educators, we surfaced concerns about using generative AI in educational settings, particularly around cultural misrepresentation, and corresponding designs for auditing tools that address these issues. We find that educators envision tools grounded in specific Hawaiian cultural values and practices, such as tracing the genealogy of knowledge in source materials. Building on these insights, we conceptualize AI auditing as a community-oriented process rather than the work of isolated individuals, and discuss implications for designing auditing tools.

We often treat social media as a lens onto society. How might that lens distort the popularity of political and social viewpoints? We examine discrepancies between publicly posted and privately surveyed opinions within communities, contributing a measurement of the “spiral of silence” theory; the theory posits people are less likely to voice opinions when they believe they hold minority views, creating a reinforcing cycle where these opinions are expressed less. We surveyed members of politically-oriented Reddit communities about their willingness to post on contentious topics, yielding 439 responses across twelve subreddits. 72.1% of participants who perceive themselves in the minority remain silent and are half as likely to post compared to those who believe their opinion is in the majority. Community design factors, such as perceived diversity, are associated with less self-silencing. We provide recommendations for counteracting self-silencing at the community level (e.g., positive reinforcement, more transparent moderation). Overall, these results reveal gaps between online discourse and broader public opinion.

Voice assistants, such as Siri and Google Assistant, typically model audio and text separately, resulting in lost speech information and increased complexity. Recent efforts to address this with end-to-end Speech Large Language Models (speech-in, text-out) trained with supervised finetuning (SFT) have led to models “forgetting” capabilities from text-only LLMs. Our work proposes an alternative paradigm for training Speech LLMs without instruction data, using the response of a text-only LLM to transcripts as self-supervision. Importantly, this process can be performed without annotated responses. We show that our Distilled Voice Assistant (DiVA) generalizes to Spoken Question Answering, Classification, and Translation. Furthermore, DiVA better matches user preferences, achieving a 72% win rate compared with state-of-the-art models like Qwen 2 Audio, despite using >100x less training compute.

Minzhi Li, William Held, Michael J. Ryan, Kunat Pipatanakul, Potsawee Manakul, Hao Zhu, Diyi Yang. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Michael J. Ryan, Omar Shaikh, Aditri Bhagirath, Daniel Frees, William Held, Diyi Yang. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Autonomous agents powered by large vision and language models (VLM) have demonstrated significant potential in completing daily computer tasks, such as browsing the web to book travel and operating desktop software, which requires agents to understand these interfaces. Despite such visual inputs becoming more integrated into agentic applications, what types of risks and attacks exist around them still remain unclear. In this work, we demonstrate that VLM agents can be easily attacked by a set of carefully designed adversarial pop-ups, which human users would typically recognize and ignore. This distraction leads agents to click these pop-ups instead of performing their tasks as usual. Integrating these pop-ups into existing agent testing environments like OSWorld and VisualWebArena leads to an attack success rate (the frequency of the agent clicking the pop-ups) of 86% on average and decreases the task success rate by 47%. Basic defense techniques, such as asking the agent to ignore pop-ups or including an advertisement notice, are ineffective against the attack. Code is available at this link.

Unlearning evaluation has traditionally followed the retrieval paradigm, where adversaries attempt to extract residual knowledge of an unlearning target by issuing queries to a language model.However, the absence of retrievable knowledge does not necessarily prevent an adversary from inferring which targets have been intentionally unlearned in the post-training optimization.Such inferences can still pose significant privacy risks, as they may reveal the sensitive data in the model's training set and the internal policies of model creators.To quantify such privacy risks, we propose a new evaluation framework Forensic Unlearning Membership Attacks (FUMA), drawing on principles from membership inference attacks.FUMA assesses whether unlearning leaves behind detectable artifacts that can be exploited to infer membership in the forget set.Specifically, we evaluate four major optimization-based unlearning methods on 258 models across diverse unlearned settings and show that examples in the forget set can be identified with up to 99% accuracy.This highlights privacy risks not covered in existing retrieval-based benchmarks.We conclude by discussing recommendations to mitigate these vulnerabilities.

Anjiang Wei, Jiannan Cao, Ran Li, Hongyu Chen, Yuhui Zhang, Ziheng Wang, Yuan Liu, Thiago S. F. X. Teixeira, Diyi Yang, Ke Wang, Alex Aiken. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025.

To serve global users safely and productively, LLMs need culture-specific knowledge that might not be learned during pre-training.How do we find knowledge that is (1) salient to ingroup users, but (2) unknown to LLMs?The most common solutions are single-initiative: either researchers define challenging questions that users passively answer (traditional annotation), or users actively produce data that researchers structure as benchmarks (knowledge extraction).The process would benefit from mixed-initiative collaboration, where users guide the process to meaningfully reflect their cultures, and LLMs steer the process to meet the researcher's goals.We propose CULTURE CARTOGRAPHY as a methodology that operationalizes this mixed-initiative vision.Here, an LLM initializes annotation with questions for which it has low-confidence answers, making explicit both its prior knowledge and the gaps therein.This allows a human respondent to fill these gaps and steer the model towards salient topics through direct edits.We implement CULTURE CARTOGRAPHY as a tool called CULTURE EXPLORER.Compared to a baseline where humans answer LLMproposed questions, we find that CULTURE EX-PLORER more effectively produces knowledge that strong models like DeepSeek R1, Llama-4 and GPT-4o are missing, even with web search.Fine-tuning on this data boosts the accuracy of Llama models by up to 19.2% on related culture benchmarks.

Language models are aligned to emulate the collective voice of many, resulting in outputs that align with no one in particular. Steering LLMs away from generic output is possible through supervised finetuning or RLHF, but requires prohibitively large datasets for new ad-hoc tasks. We argue that it is instead possible to align an LLM to a specific setting by leveraging a very small number ($<10$) of demonstrations as feedback. Our method, Demonstration ITerated Task Optimization (DITTO), directly aligns language model outputs to a user's demonstrated behaviors. Derived using ideas from online imitation learning, DITTO cheaply generates online comparison data by treating users' demonstrations as preferred over output from the LLM and its intermediate checkpoints. We evaluate DITTO's ability to learn fine-grained style and task alignment across domains such as news articles, emails, and blog posts. Additionally, we conduct a user study soliciting a range of demonstrations from participants ($N=16$). Across our benchmarks and user study, we find that win-rates for DITTO outperform few-shot prompting, supervised fine-tuning, and other self-play methods by an average of 19\% points. By using demonstrations as feedback directly, DITTO offers a novel method for effective customization of LLMs.

Recent advancements in large language models (LLMs) have sparked optimism about their potential to accelerate scientific discovery, with a growing number of works proposing research agents that autonomously generate and validate new ideas. Despite this, no evaluations have shown that LLM systems can take the very first step of producing novel, expert-level ideas, let alone perform the entire research process. We address this by establishing an experimental design that evaluates research idea generation while controlling for confounders and performs the first comparison between expert NLP researchers and an LLM ideation agent. By recruiting over 100 NLP researchers to write novel ideas and blind reviews of both LLM and human ideas, we obtain the first statistically significant conclusion on current LLM capabilities for research ideation: we find LLM-generated ideas are judged as more novel (p < 0.05) than human expert ideas while being judged slightly weaker on feasibility. Studying our agent baselines closely, we identify open problems in building and evaluating research agents, including failures of LLM self-evaluation and their lack of diversity in generation.

Autonomous systems for software engineering are now capable of fixing bugs and developing features. These systems are commonly evaluated on SWE-bench (Jimenez et al., 2024a), which assesses their ability to solve software issues from GitHub repositories. However, SWE-bench uses only Python repositories, with problem statements presented predominantly as text and lacking visual elements such as images. This limited coverage motivates our inquiry into how existing systems might perform on unrepresented software engineering domains (e.g., front-end, game development, DevOps), which use different programming languages and paradigms. Therefore, we propose SWE-bench Multimodal (SWE-bench M), to evaluate systems on their ability to fix bugs in visual, user-facing JavaScript software. SWE-bench M features 617 task instances collected from 17 JavaScript libraries used for web interface design, diagramming, data visualization, syntax highlighting, and interactive mapping. Each SWE-bench M task instance contains at least one image in its problem statement or unit tests. Our analysis finds that top-performing SWE-bench systems struggle with SWE-bench M, revealing limitations in visual problem-solving and cross-language generalization. Lastly, we show that SWE-agent’s flexible language-agnostic features enable it to substantially outperform alternatives on SWE-bench M, resolving 12% of task instances compared to 6% for the next best system.

Preferences within a group of people are not uniform but follow a distribution. While existing alignment methods like Direct Preference Optimization (DPO) attempt to steer models to reflect human preferences, they struggle to capture the distributional pluralistic preferences within a group. These methods often skew toward dominant preferences, overlooking the diversity of opinions, especially when conflicting preferences arise. To address this issue, we propose Group Distributional Preference Optimization (GDPO), a novel framework that aligns language models with the distribution of preferences within a group by incorporating the concept of beliefs that shape individual preferences. GDPO calibrates a language model using statistical estimation of the group's belief distribution and aligns the model with belief-conditioned preferences, offering a more inclusive alignment framework than traditional methods. In experiments using both synthetic controllable opinion generation and real-world movie review datasets, we show that DPO fails to align with the targeted belief distributions, while GDPO consistently reduces this alignment gap during training. Additionally, our evaluation metrics demonstrate that GDPO outperforms existing approaches in aligning with group distributional preferences, marking a significant advance in pluralistic alignment.

Interpretability research now offers a variety of techniques for identifying abstract internal mechanisms in neural networks. Can such techniques be used to predict how models will behave on out-of-distribution examples? In this work, we provide a positive answer to this question. Through a diverse set of language modeling tasks—including symbol manipulation, knowledge retrieval, and instruction following—we show that the most robust features for correctness prediction are those that play a distinctive causal role in the model’s behavior. Specifically, we propose two methods that leverage causal mechanisms to predict the correctness of model outputs: counterfactual simulation (checking whether key causal variables are realized) and value probing (using the values of those variables to make predictions). Both achieve high AUC-ROC in distribution and outperform methods that rely on causal-agnostic features in out-of-distribution settings, where predicting model behaviors is more crucial. Our work thus highlights a novel and significant application for internal causal analysis of language models.

Ryan Li, Yanzhe Zhang, Diyi Yang. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2025.

Chenglei Si, Yanzhe Zhang, Ryan Li, Zhengyuan Yang, Ruibo Liu, Diyi Yang. Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2025.

Suppose Alice trains an open-weight language model and Bob uses a blackbox derivative of Alice’s model to produce text. Can Alice prove that Bob is using her model, either by querying Bob’s derivative model (query setting) or from the text alone ( observational setting)? We formulate this question as an independence testing problem—in which the null hypothesis is that Bob’s model or text is independent of Alice’s randomized training run—and investigate it through the lens of palimpsestic memorization in language models: models are more likely to memorize data seen later in training, so we can test whether Bob is using Alice’s model using test statistics that capture correlation between Bob’s model or text and the ordering of training examples in Alice’s training run. If Alice has randomly shuffled her training data, then any significant correlation amounts to exactly quantifiable statistical evidence against the null hypothesis, regardless of the composition of Alice’s training data. In the query setting, we directly estimate (via prompting) the likelihood Bob’s model gives to Alice’s training examples and their training order; we correlate the likelihoods of over 40 fine-tunes of various Pythia and OLMo base models ranging from 1B to 12B parameters with the base model’s training data order, achieving a p-value on the order of at most $1 \times 10^{-8}$ in all but six cases. In the observational setting, we try two approaches based on estimating 1) the likelihood of Bob’s text overlapping with spans of Alice’s training examples and 2) the likelihood of Bob’s text with respect to different versions of Alice’s model we obtain by repeating the last phase (e.g., 1%) of her training run on reshuffled data. The second approach can reliably distinguish Bob’s text from as little as a few hundred tokens; the first does not involve any retraining but requires many more tokens (several hundred thousand) to achieve high power.

Recent advances in general-purpose AI underscore the urgent need to align AI systems with human goals and values. Yet, the lack of a clear, shared understanding of what constitutes "alignment" limits meaningful progress and cross-disciplinary collaboration. In this position paper, we argue that the research community should explicitly define and critically reflect on "alignment" to account for the bidirectional and dynamic relationship between humans and AI. Through a systematic review of over 400 papers spanning HCI, NLP, ML, and more, we examine how alignment is currently defined and operationalized. Building on this analysis, we introduce the Bidirectional Human-AI Alignment framework, which not only incorporates traditional efforts to align AI with human values but also introduces the critical, underexplored dimension of aligning humans with AI – supporting cognitive, behavioral, and societal adaptation to rapidly advancing AI technologies. Our findings reveal significant gaps in current literature, especially in long-term interaction design, human value modeling, and mutual understanding. We conclude with three central challenges and actionable recommendations to guide future research toward more nuanced, reciprocal, and human-AI alignment approaches.

As large language models (LLMs) increasingly serve as close collaborators for humans, it is crucial that they express their reasoning in ways that humans can understand and learn from. However, this capability remains relatively less understood and under-evaluated. To address this, we introduce a conceptual framework for such Human-AI knowledge transfer capabilities and conduct the first large-scale user study (N=118) explicitly designed to measure it. In our two-phase setup, humans first ideate with an LLM on problem-solving strategies, then independently implement solutions, isolating the influence of model reasoning on human understanding. Our findings reveal that while model benchmark performance correlates with collaborative outcomes, this relationship is notably inconsistent with significant outliers, highlighting that knowledge transfer is a distinct capability requiring dedicated optimization. Our analysis uncovers behavioral and strategic factors that mediate successful knowledge transfer, and we release our code, dataset, and evaluation framework to support future work on communicatively aligned models.

Vision-language models have demonstrated impressive capabilities as computer-use agents (CUAs) capable of automating diverse computer tasks. As their commercial potential grows, critical details of the most capable CUA systems remain closed. As these agents will increasingly mediate digital interactions and execute consequential decisions on our behalf, the research community needs access to open CUA frameworks to study their capabilities, limitations, and risks. To bridge this gap, we propose OpenCUA, a comprehensive open-source framework for scaling CUA data and foundation models. Our framework consists of: (1) an annotation infrastructure that seamlessly captures human computer-use demonstrations; (2) AgentNet, the first large-scale computer-use task dataset spanning 3 operating systems and 200+ applications and websites; (3) a scalable pipeline that transforms demonstrations into state–action pairs with reflective long Chain-of-Thought reasoning that sustain robust performance gains as data scales. Our end-to-end agent models demonstrate strong performance across CUA benchmarks. In particular, OpenCUA-72B achieves an average success rate of 45.0% on OSWorld‑Verified, establishing a new state-of-the-art (SOTA) among open-source models. Further analysis confirms that our approach generalizes well across domains and benefits significantly from increased test-time computation. We release our annotation tool, datasets, code, and models to build open foundations for further CUA research.

Despite recent progress in Language Models (LMs) for software engineering, collecting training data remains a significant pain point.Existing datasets are small, with at most 1,000s of training instances from 11 or fewer GitHub repositories.The procedures to curate such datasets are often complex, necessitating hundreds of hours of human labor; companion execution environments also take up several terabytes of storage, severely limiting their scalability and usability.To address this pain point, we introduce SWE-smith, a novel pipeline for generating software engineering training data at scale.Given any Python codebase, SWE-smith constructs a corresponding execution environment, then automatically synthesizes 100s to 1,000s of task instances that break existing test(s) in the codebase.Using SWE-smith, we create a dataset of 50k instances sourced from 128 GitHub repositories, an order of magnitude larger than all previous works.We train SWE-agent-LM-32B, achieving 40.2% Pass@1 resolve rate on the SWE-bench Verified benchmark, state of the art among open source models.We open source SWE-smith (collection procedure, task instances, trajectories, models) to lower the barrier of entry for research in LM systems for automated software engineering.All assets available at \url{https://swesmith.com}.

Despite the growing integration of retrieval-enabled AI agents into society, their safety and ethical behavior remain inadequately understood. In particular, the growing integration of LLMs and AI agents with external information sources and real-world environments raises critical questions about how they engage with and are influenced by these external data sources and interactive contexts. This study investigates how expanding retrieval access—from no external sources to Wikipedia-based retrieval and open web search—affects model reliability, bias propagation, and harmful content generation. Through extensive benchmarking of censored and uncensored LLMs and AI Agents, our findings reveal a consistent degradation in refusal rates, bias sensitivity, and harmfulness safeguards as models gain broader access to external sources, culminating in a phenomenon we term safety degradation. Notably, retrieval-enabled agents built on aligned LLMs often behave more unsafely than uncensored models without retrieval. This effect persists even under strong retrieval accuracy and prompt-based mitigation, suggesting that the mere presence of retrieved content reshapes model behavior in structurally unsafe ways. These findings underscore the need for robust mitigation strategies to ensure fairness and reliability in retrieval-enabled and increasingly autonomous AI systems.

Millions of users come to online peer counseling platforms to seek support. However, studies show that online peer support groups are not always as effective as expected, largely due to users' negative experiences with unhelpful counselors. Peer counselors are key to the success of online peer counseling platforms, but most often do not receive appropriate training. Hence, we introduce CARE: an AI-based tool to empower and train peer counselors through practice and feedback. Concretely, CARE helps diagnose which counseling strategies are needed in a given situation and suggests example responses to counselors during their practice sessions. Building upon the Motivational Interviewing framework, CARE utilizes large-scale counseling conversation data with text generation techniques to enable these functionalities. We demonstrate the efficacy of CARE by performing quantitative evaluations and qualitative user studies through simulated chats and semi-structured interviews, finding that CARE especially helps novice counselors in challenging situations. The code is available at https://github.com/SALT-NLP/CARE.

What challenges do volunteers providing peer support in online mental health platforms (OMHPs) face in operating and growing their communities? How could the HCI community develop human-AI systems to help? Recent work on online peer counseling has led to the development of novel AI tools for conversational interaction, but it remains unknown how such technology can fit into broader practices that include extratherapeutic tasks. In this research, we conducted interviews and design exercises with seventeen peer counselors from 7 Cups of Tea, a large online therapy and counseling platform, to design tools --- AI or not --- that resolve challenges that arise from day-to-day community practices. Participant responses suggest three classes of tools that could improve online peer counseling: real-time decision support, productivity, and management and training. Investigation of design motivations surfaced four practice-based challenges including chat interface limitations, difficulties in support seeker management, fragmented contexts of practice, and lack of visibility due to privacy concerns. Based on counselors' discussion of benefits and risks associated with AI features in the tools they designed, we offer suggestions for research on AI tools embedded within peer counseling practices, and connect our findings with broader implications about online peer counseling as a form of volunteer-based mental health practice.

We present MIDDAG, an intuitive, interactive system that visualizes the information propagation paths on social media triggered by COVID-19-related news articles accompanied by comprehensive insights including user/community susceptibility level, as well as events and popular opinions raised by the crowd while propagating the information. Besides discovering information flow patterns among users, we construct communities among users and develop the propagation forecasting capability, enabling tracing and understanding of how information is disseminated at a higher level. A demo video and more are available at https://info-pathways.github.io.

Most traditional AI safety research views models as machines and centers on algorithmfocused attacks developed by security experts.As large language models (LLMs) become increasingly common and competent, non-expert users can also impose risks during daily interactions.Observing this, we shift the perspective, by treating LLMs as human-like communicators to examine the interplay between everyday language interaction and AI safety.Specifically, we study how to persuade LLMs to jailbreak them.First, we propose a persuasion taxonomy derived from decades of social science research.Then, we apply the taxonomy to automatically generate persuasive adversarial prompts (PAP) to jailbreak LLMs.Results show that persuasion significantly increases the jailbreak risk across all risk categories: PAP consistently achieves an attack success rate of over 92% on Llama-2-7b-Chat, GPT-3.5, and GPT-4 in 10 trials, surpassing recent algorithm-focused attacks.On the defense side, we explore various mechanisms against PAP, find a significant gap in existing defenses, and advocate for more fundamental solutions for AI safety 1 .

Realistic practice and tailored feedback are key processes for training peer counselors with clinical skills.However, existing mechanisms of providing feedback largely rely on human supervision.Peer counselors often lack mechanisms to receive detailed feedback from experienced mentors, making it difficult for them to support the large number of people with mental health issues who use peer counseling.Our work aims to leverage large language models to provide contextualized and multi-level feedback to empower peer counselors, especially novices, at scale.To achieve this, we co-design with a group of senior psychotherapy supervisors to develop a multi-level feedback taxonomy, and then construct a publicly available dataset with comprehensive feedback annotations of 400 emotional support conversations.We further design a self-improvement method on top of large language models to enhance the automatic generation of feedback.Via qualitative and quantitative evaluation with domain experts, we demonstrate that our method minimizes the risk of potentially harmful and lowquality feedback generation which is desirable in such high-stakes scenarios.

Julia Kruk, Michela Marchini, Rijul Magu, Caleb Ziems, David Muchlinski, Diyi Yang. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

Before being deployed for user-facing applications, developers align Large Language Models (LLMs) to user preferences through a variety of procedures, such as Reinforcement Learning From Human Feedback (RLHF) and Direct Preference Optimization (DPO).Current evaluations of these procedures focus on benchmarks of instruction following, reasoning, and truthfulness.However, human preferences are not universal, and aligning to specific preference sets may have unintended effects.We explore how alignment impacts performance along three axes of global representation: English dialects, multilingualism, and opinions from and about countries worldwide.Our results show that current alignment procedures create disparities between English dialects and global opinions.We find alignment improves capabilities in several languages.We conclude by discussing design decisions that led to these unintended impacts and recommendations for more equitable preference tuning.We make our code and data publicly available on Github 1 .

Image editing tools have a plethora of commercial and creative applications — content-creation, digital photography, advertisements, graphic design, and development of educational media. The shortcomings of image editing software include difficulty of use and, for AI-based software, reliance on single image editing models, which often poses the dilemma of a tradeoff between image editing quality and user-friendliness. While the performances of individual image editing models have improved with their evolution over time, these singular models are often specialized on specific image editing tasks. In this work, we introduce HILITE, an open-source interactive image editing platform with a human-in-the-loop design that combines six diffusion-based image editing models. For one, HILITE’s accessible and easily-understandable user interface provides a straightforward user workflow from image input and prompt entry to selection of desired output. Secondly, the combination of several models with diverse specializations in turn allows HILITE to generalize on a wide variety of image editing tasks, essentially creating a "one-stop shop" for image editing. Third, HILITE iteratively takes user feedback, which both enhances the user experience and enables collection of crowd-sourced data for image editing. HILITE outperforms two major image editing softwares, OpenAI’s DALL•E 3 and Google’s Imagen 3, across two widely-user quantitative metrics for image editing evaluation. Considering the growing demand for readily-available and high-performing image editing tools, HILITE provides a novel platform design with multifaceted use cases in both business and academia. The platform can be found at https://platform.opennlplabs.org/ or https://platform-deployment.vercel.app/.

Interpersonal conflict is an uncomfortable but unavoidable fact of life. Navigating conflict successfully is a skill—one that can be learned through deliberate practice—but few have access to effective training or feedback. To expand this access, we introduce Rehearsal, a system that allows users to rehearse conflicts with a believable simulated interlocutor, explore counterfactual “what if?” scenarios to identify alternative conversational paths, and learn through feedback on how and when to apply specific conflict strategies. Users can utilize Rehearsal to practice handling a variety of predefined conflict scenarios, from office disputes to relationship issues, or they can choose to create their own setting. To enable Rehearsal, we develop IRP prompting, a method of conditioning output of a large language model on the influential Interest-Rights-Power (IRP) theory from conflict resolution. Rehearsal uses IRP to generate utterances grounded in conflict resolution theory, guiding users towards counterfactual conflict resolution strategies that help de-escalate difficult conversations. In a between-subjects evaluation, 40 participants engaged in an actual conflict with a confederate after training. Compared to a control group with lecture material covering the same IRP theory, participants with simulated training from Rehearsal significantly improved their performance in the unaided conflict: they reduced their use of escalating competitive strategies by an average of 67%, while doubling their use of cooperative strategies. Overall, Rehearsal highlights the potential effectiveness of language models as tools for learning and practicing interpersonal skills.

Modern language models often exhibit powerful but brittle behavior, leading to the development of larger and more diverse benchmarks to reliably assess their behavior.Here, we suggest that model performance can be benchmarked and elucidated with much smaller evaluation sets.We first show that in six popular language classification benchmarks, model confidence in the correct class on many pairs of points is strongly correlated across models.We build upon this phenomenon to propose Anchor Point Selection, a technique to select small subsets of datasets that capture model behavior across the entire dataset.Anchor points reliably rank models: across 87 diverse language model-prompt pairs, evaluating models using 1-30 anchor points outperforms uniform sampling and other baselines at accurately ranking models.Moreover, just a dozen anchor points can be used to estimate model per-class predictions on all other points in a dataset with low error, sufficient for gauging where the model is likely to fail.Lastly, we present Anchor Point Maps for visualizing these insights and facilitating comparisons of the performance of different models on various regions within the dataset distribution.

Large Language Models (LLMs) frequently memorize long sequences verbatim, often with serious legal and privacy implications.Much prior work has studied such verbatim memorization using observational data.To complement such work, we develop a framework to study verbatim memorization in a controlled setting by continuing pre-training from Pythia checkpoints with injected sequences.We find that (1) non-trivial amounts of repetition are necessary for verbatim memorization to happen; (2) later (and presumably better) checkpoints are more likely to verbatim memorize sequences, even for out-of-distribution sequences; (3) the generation of memorized sequences is triggered by distributed model states that encode high-level features and makes important use of general language modeling capabilities.Guided by these insights, we develop stress tests to evaluate unlearning methods and find they often fail to remove the verbatim memorized information, while also degrading the LM.Overall, these findings challenge the hypothesis that verbatim memorization stems from specific model weights or mechanisms.Rather, verbatim memorization is intertwined with the LM's general capabilities and thus will be very difficult to isolate and suppress without degrading model quality. The Original Trigger PrefixMr and Mrs Dursley, of number four, Privet Drive, were proud to say that they were perfectly normal, thank you very much Trigger Prefixes with Similar High-level FeaturesMrs and Mr Dursley, of number four, Privet Drive, were proud to say that they were perfectly normal, thank you very much The Dursley family, of number four, Privet Drive, were proud to say that they were perfectly normal, thank you very much Mr and Mrs Weasley, residing at four Privet Drive, were proud to say they were perfectly normal, thank you very much Mr and Mrs Slytherin, of number twenty-one, Privet Drive, were proud to say that they were perfectly normal, thank you very much Mr and Mrs Dursley, of #4, Privet Drive, were proud to say that they were perfectly normal, thank you very much Mr and Mrs Dursley, of number ten, Privet Drive, were proud to say that they were perfectly normal, thank you very much Mr and Mrs Dursley, of Privet Drive, were proud to say that they were perfectly normal, thank you very much Mr and Mrs Dursley, of number four, Oak Street, were proud to say that they were perfectly normal, thank you very much Mr and Mrs Dursley, residing at four Privet Drive, were delighted to assert they were perfectly normal, thank you very much The Dursley family, of number four, Privet Drive, were pleased to declare that they were perfectly normal, thank you very much Non-Trigger Prefixes with Similar or Different High-level FeaturesMr and Mrs Kingsley, of number four, Privet Drive, were proud to say that they were the proud parents of a bouncing baby boy.Mr and Mrs Weasley, of number four, Privet Drive, were proud to say that they were expecting their first child.Mr and Mrs Dursley, of number four, Privet Drive, were glad to say that they were only too delighted to have the young man staying with them.

Susceptibility to misinformation describes the degree of belief in unverifiable claims, a latent aspect of individuals' mental processes that is not observable.Existing susceptibility studies heavily rely on self-reported beliefs, which can be subject to bias, expensive to collect, and challenging to scale for downstream applications.To address these limitations, in this work, we propose a computational approach to efficiently model users' latent susceptibility levels.As shown in previous work, susceptibility is influenced by various factors (e.g., demographic factors, political ideology), and directly influences people's reposting behavior on social media.To represent the underlying mental process, our susceptibility modeling incorporates these factors as inputs, guided by the supervision of people's sharing behavior.Using COVID-19 as a testbed, our experiments demonstrate a significant alignment between the susceptibility scores estimated by our computational modeling and human judgments, confirming the effectiveness of this latent modeling approach.Furthermore, we apply our model to annotate susceptibility scores on a large-scale dataset and analyze the relationships between susceptibility with various factors.Our analysis reveals that political leanings and other psychological factors exhibit varying degrees of association with susceptibility to COVID-19 misinformation, and shows that susceptibility is unevenly distributed across different professional and geographical backgrounds. 1

Recent works leverage LLMs to roleplay realistic social scenarios, aiding novices in practicing their social skills.However, simulating sensitive interactions, such as in the domain of mental health, is challenging.Privacy concerns restrict data access, and collecting expert feedback, although vital, is laborious.To address this, we develop Roleplay-doh, a novel human-LLM collaboration pipeline that elicits qualitative feedback from a domain-expert, which is transformed into a set of principles, or natural language rules, that govern an LLMprompted roleplay.We apply this pipeline to enable senior mental health supporters to create customized AI patients as simulated practice partners for novice counselors.After uncovering issues with basic GPT-4 simulations not adhering to expert-defined principles, we also introduce a novel principle-adherence prompting pipeline which shows a 30% improvement in response quality and principle following for the downstream task.Through a user study with 25 counseling experts, we demonstrate that the pipeline makes it easy and effective to create AI patients that more faithfully resemble real patients, as judged by both creators and third-party counselors.We provide access to the code and data on our project website 1 .

The goal of social alignment for AI systems is to make sure these models can conduct themselves appropriately following social values. Unlike humans who establish a consensus on value judgments through social interaction, current language models (LMs) are trained to rigidly recite the corpus in social isolation, which causes poor generalization in unfamiliar cases and the lack of robustness under adversarial attacks. In this work, we introduce a new training paradigm that enables LMs to learn from simulated social interactions. Compared with existing methods, our method is much more scalable and efficient, and shows superior performance in alignment benchmarks and human evaluation.

Large language models (LLMs) have achieved remarkable performance in various evaluation benchmarks. However, concerns are raised about potential data contamination in their considerable volume of training corpus. Moreover, the static nature and fixed complexity of current benchmarks may inadequately gauge the advancing capabilities of LLMs. In this paper, we introduce DyVal, a general and flexible protocol for dynamic evaluation of LLMs. Based on our framework, we build graph-informed DyVal by leveraging the structural advantage of directed acyclic graphs to dynamically generate evaluation samples with controllable complexities. DyVal generates challenging evaluation sets on reasoning tasks including mathematics, logical reasoning, and algorithm problems. We evaluate various LLMs ranging from Flan-T5-large to GPT-3.5-Turbo and GPT-4. Experiments show that LLMs perform worse in DyVal-generated evaluation samples with different complexities, highlighting the significance of dynamic evaluation. We also analyze the failure cases and results of different prompting methods. Moreover, DyVal-generated samples are not only evaluation sets, but also helpful data for fine-tuning to improve the performance of LLMs on existing benchmarks. We hope that DyVal can shed light on future evaluation research of LLMs. Code is available at: https://github.com/microsoft/promptbench.

Independent evaluation and red teaming are critical for identifying the risks posed by generative AI systems. However, the terms of service and enforcement strategies used by prominent AI companies to deter model misuse have disincentives on good faith safety evaluations. This causes some researchers to fear that conducting such research or releasing their findings will result in account suspensions or legal reprisal. Although some companies offer researcher access programs, they are an inadequate substitute for independent research access, as they have limited community representation, receive inadequate funding, and lack independence from corporate incentives. We propose that major generative AI developers commit to providing a legal and technical safe harbor, protecting public interest safety research and removing the threat of account suspensions or legal reprisal. These proposals emerged from our collective experience conducting safety, privacy, and trustworthiness research on generative AI systems, where norms and incentives could be better aligned with public interests, without exacerbating model misuse. We believe these commitments are a necessary step towards more inclusive and unimpeded community efforts to tackle the risks of generative AI.

Growing concerns have been raised about the detrimental effects of uncivil comments on the web towards democracy. However, there is still a lack of understanding about online incivility's nuanced and complicated nature and its impact on conversation development and user behaviors. This work aims to fill that research gap by modeling incivility and its relationship to political discussions. We develop a comprehensive and fine-grained taxonomy that characterizes incivility with vulgarity, name-calling (inter-personal and third-party attacks), aspersion, and stereotypes, and then apply the framework to quantify the level of each incivility category in over 40 million comments from Reddit. Using large-scale quantitative analysis, we investigate the types of interactions and contexts in which incivility is more likely to occur, model how incivility shapes subsequent conversations, and examine user engagement patterns and behavioral changes after exposure to incivility. Our findings show that conversations that start out uncivil tend to become more uncivil in responses, and exposure to different incivility categories has differing effects on community members' engagement. We conclude with the implications of our research in assisting the design and moderation of online political communities.

Omar Shaikh, Kristina Gligoric, Ashna Khetan, Matthias Gerstgrasser, Diyi Yang, Dan Jurafsky. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2024.

Text-to-image diffusion models have impactful applications in art, design, and entertainment, yet these technologies also pose significant risks by enabling the creation and dissemination of misinformation. Although recent advancements have produced AI-generated image detectors that claim robustness against various augmentations, their true effectiveness remains uncertain. Do these detectors reliably identify images with different levels of augmentation? Are they biased toward specific scenes or data distributions? To investigate, we introduce **Semi-Truths**, featuring $27,600$ real images, $223,400$ masks, and $1, 329, 155$ AI-augmented images that feature targeted and localized perturbations produced using diverse augmentation techniques, diffusion models, and data distributions. Each augmented image is accompanied by metadata for standardized and targeted evaluation of detector robustness. Our findings suggest that state-of-the-art detectors exhibit varying sensitivities to the types and degrees of perturbations, data distributions, and augmentation methods used, offering new insights into their performance and limitations. The code for the augmentation and evaluation pipeline is available at https://github.com/J-Kruk/SemiTruths.

As language models (LMs) are widely utilized in personalized communication scenarios (e.g., sending emails, writing social media posts) and endowed with a certain level of agency, ensuring they act in accordance with the contextual privacy norms becomes increasingly critical. However, quantifying the privacy norm awareness of LMs and the emerging privacy risk in LM-mediated communication is challenging due to (1) the contextual and long-tailed nature of privacy-sensitive cases, and (2) the lack of evaluation approaches that capture realistic application scenarios. To address these challenges, we propose PrivacyLens, a novel framework designed to extend privacy-sensitive seeds into expressive vignettes and further into agent trajectories, enabling multi-level evaluation of privacy leakage in LM agents' actions. We instantiate PrivacyLens with a collection of privacy norms grounded in privacy literature and crowdsourced seeds. Using this dataset, we reveal a discrepancy between LM performance in answering probing questions and their actual behavior when executing user instructions in an agent setup. State-of-the-art LMs, like GPT-4 and Llama-3-70B, leak sensitive information in 25.68% and 38.69% of cases, even when prompted with privacy-enhancing instructions. We also demonstrate the dynamic nature of PrivacyLens by extending each seed into multiple trajectories to red-team LM privacy leakage risk. Dataset and code are available at https://github.com/SALT-NLP/PrivacyLens.

The current paradigm of evaluating Large Language Models (LLMs) through static benchmarks comes with significant limitations, such as vulnerability to data contamination and a lack of adaptability to the evolving capabilities of LLMs. Therefore, evaluation methods that can adapt and generate evaluation data with controlled complexity are urgently needed. In this work, we introduce Dynamic Evaluation of LLMs via Adaptive Reasoning Graph Evolvement (DARG) to dynamically extend current benchmarks with controlled complexity and diversity. Specifically, we first extract the reasoning graphs of data points in current benchmarks and then perturb the reasoning graphs to generate novel testing data. Such newly generated test samples can have different levels of complexity while maintaining linguistic diversity similar to the original benchmarks. We further use a code-augmented LLM to ensure the label correctness of newly generated data. We apply our DARG framework to diverse reasoning tasks in four domains with 15 state-of-the-art LLMs. Experimental results show that almost all LLMs experience a performance decrease with increased complexity and certain LLMs exhibit significant drops. Additionally, we find that LLMs exhibit more biases when being evaluated via the data generated by DARG with higher complexity levels. These observations provide useful insights into how to dynamically and adaptively evaluate LLMs.

Abstract Large language models (LLMs) are capable of successfully performing many language processing tasks zero-shot (without training data). If zero-shot LLMs can also reliably classify and explain social phenomena like persuasiveness and political ideology, then LLMs could augment the computational social science (CSS) pipeline in important ways. This work provides a road map for using LLMs as CSS tools. Towards this end, we contribute a set of prompting best practices and an extensive evaluation pipeline to measure the zero-shot performance of 13 language models on 25 representative English CSS benchmarks. On taxonomic labeling tasks (classification), LLMs fail to outperform the best fine-tuned models but still achieve fair levels of agreement with humans. On free-form coding tasks (generation), LLMs produce explanations that often exceed the quality of crowdworkers’ gold references. We conclude that the performance of today’s LLMs can augment the CSS research pipeline in two ways: (1) serving as zero-shot data annotators on human annotation teams, and (2) bootstrapping challenging creative generation tasks (e.g., explaining the underlying attributes of a text). In summary, LLMs are posed to meaningfully participate in social science analysis in partnership with humans.

When discussing difficult topics online, is it common to meaningfully engage with people from diverse perspectives? Why or why not? Could features of the online environment be redesigned to encourage civil conversation across difference? To investigate these questions, we need to explore them in a particular context. In this paper, we study discussions of gun policy on Reddit, with the overarching goal of developing insights into the potential of the internet to support understanding across difference. We use two methods: a clustering analysis of Reddit posts to contribute insights about what people discuss, and an interview study of twenty Reddit users to help us understand why certain kinds of conversation take place and others don't. We find that the discussion of gun politics falls into three groups: conservative pro-gun, liberal pro-gun, and liberal anti-gun. Each type of group has its own characteristic topics. While our subjects state that they would be willing to engage with others across the ideological divide, in practice they rarely do. Subjects are siloed into like-minded subreddits through a two-pronged effect, where they are simultaneously pushed away from opposing-view communities while actively seeking belonging in like-minded ones. Another contributing factor is Reddit's "karma" mechanism: fear of being downvoted and losing karma points and social approval of peers causes our subjects to hesitate to say anything in conflict with group norms. The pseudonymous nature of discussion on Reddit plays a complex role, with some subjects finding it freeing and others fearing reprisal from others not bound by face-to-face norms of politeness. Our subjects believe that content moderation can help ameliorate these issues; however, our findings suggest that moderators need different tools to do so effectively. We conclude by suggesting platform design changes that might increase discussion across difference.

Online mental health support communities, in which volunteer counselors provide accessible mental and emotional health support, have grown in recent years. Despite millions of people using these platforms, the clinical effectiveness of these communities on mental health symptoms remains unknown. Although volunteers receive some training on the therapeutic skills proven effective in face-to-face environments, such as active listening and motivational interviewing, it is unclear how the usage of these skills in an online context affects people's mental health. In our work, we collaborate with one of the largest online peer support platforms and use both natural language processing and machine learning techniques to examine how one-on-one support chats on the platform affect clients' depression and anxiety symptoms. We measure how characteristics of support-providers, such as their experience on the platform and use of therapeutic skills (e.g. affirmation, showing empathy), affect support-seekers' mental health changes. Based on a propensity-score matching analysis to approximate a random-assignment experiment, results shows that online peer support chats improve both depression and anxiety symptoms with a statistically significant but relatively small effect size. Additionally, support providers' techniques such as emphasizing the autonomy of the client lead to better mental health outcomes. However, we also found that the use of some behaviors, such as persuading and providing information, are associated with worsening of mental health symptoms. Our work provides key understanding for mental health care in the online setting and designing training systems for online support providers.

William Held, Christopher Hidey, Fei Liu, Eric Zhu, Rahul Goel, Diyi Yang, Rushin Shah. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2023.

Recent abstractive conversation summarization systems generally rely on large-scale datasets with annotated summaries.However, collecting and annotating these conversations can be a time-consuming and labor-intensive task.To address this issue, in this work, we present a sub-structure level compositional data augmentation method, COMPO, for generating diverse and high-quality pairs of conversations and summaries.Specifically, COMPO first extracts conversation structures like topic splits and action triples as basic units.Then we organize these semantically meaningful conversation snippets compositionally to create new training instances.Additionally, we explore noise-tolerant settings in both self-training and joint-training paradigms to make the most of these augmented samples.Our experiments on benchmark datasets, SAMSum and Di-alogSum, show that COMPO substantially outperforms prior baseline methods by achieving a nearly 10% increase of ROUGE scores with limited data.We have publically released our code at https://github.com/ ozyyshr/Compo.

Warning: This paper contains several toxic and offensive statements.Generating a Chain of Thought (CoT) has been shown to consistently improve large language model (LLM) performance on a wide range of NLP tasks.However, prior work has mainly focused on logical reasoning tasks (e.g.arithmetic, commonsense QA); it remains unclear whether improvements hold for more diverse types of reasoning, especially in socially situated contexts.Concretely, we perform a controlled evaluation of zero-shot CoT across two socially sensitive domains: harmful questions and stereotype benchmarks.We find that zeroshot CoT reasoning in sensitive domains significantly increases a model's likelihood to produce harmful or undesirable output, with trends holding across different prompt formats and model variants.Furthermore, we show that harmful CoTs increase with model size, but decrease with improved instruction following.Our work suggests that zero-shot CoT should be used with caution on socially important tasks, especially when marginalized groups or sensitive topics are involved.

Citing papers is the primary method through which modern scientific writing discusses and builds on past work.Collectively, citing a diverse set of papers (in time and area of study) is an indicator of how widely the community is reading.Yet there is little work looking at broad temporal patterns of citation.This work, systematically and empirically examines: How far back in time do we tend to go to cite papers?How has that changed over time, and what factors correlate with this citational attention/amnesia?We chose NLP as our domain of interest, and analyzed ∼71.5K papers to show and quantify several key trends in citation.Notably, ∼62% of cited papers are from the immediate five years prior to publication, whereas only ∼17% are more than ten years old.Furthermore, we show that the median age and age diversity of cited papers was steadily increasing from 1990 to 2014, but since then the trend has reversed, and current NLP papers have an all-time low temporal citation diversity.Finally, we show that unlike the 1990s, the highly cited papers in the last decade were also papers with the least citation diversity; likely contributing to the intense (and arguably harmful) recency focus.Code, data, and a demo are available at the project homepage.1 2

We present NORMBANK, a knowledge bank of 155k situational norms.This resource is designed to ground flexible normative reasoning for interactive, assistive, and collaborative AI systems.Unlike prior commonsense resources, NORMBANK grounds each inference within a multivalent sociocultural frame, which includes the setting (e.g., restaurant), the agents' contingent roles (waiter, customer), their attributes (age, gender), and other physical, social, and cultural constraints (e.g., the temperature or the country of operation).In total, NORMBANK contains 63k unique constraints from a taxonomy that we introduce and iteratively refine here.Constraints then apply in different combinations to frame social norms.Under these manipulations, norms are non-monotonic -one can cancel an inference by updating its frame even slightly.Still, we find evidence that neural models can help reliably extend the scope and coverage of NORMBANK.We further demonstrate the utility of this resource with a series of transfer experiments.For data and code, see

Dialect differences caused by regional, social, and economic factors cause performance discrepancies for many groups of language technology users.Inclusive and equitable language technology must critically be dialect invariant, meaning that performance remains constant over dialectal shifts.Current systems often fall short of this ideal since they are designed and tested on a single dialect: Standard American English (SAE).We introduce a suite of resources for evaluating and achieving English dialect invariance.The resource is called Multi-VALUE, a controllable rule-based translation system spanning 50 English dialects and 189 unique linguistic features.Multi-VALUE maps SAE to synthetic forms of each dialect.First, we use this system to stress tests question answering, machine translation, and semantic parsing.Stress tests reveal significant performance disparities for leading models on nonstandard dialects.Second, we use this system as a data augmentation technique to improve the dialect robustness of existing systems.Finally, we partner with native speakers of Chicano and Indian English to release new goldstandard variants of the popular CoQA task.To execute the transformation code, run model checkpoints, and download both synthetic and gold-standard dialectal benchmark datasets, see http://value-nlp.org/.

Extensive research has been published on the conversational factors of effective volunteer peer counseling on online mental health platforms (OMHPs). However, studies differ in how they define and measure success outcomes, with most prior work examining only a single success metric. In this work, we model the relationship between previously reported linguistic predictors of effective counseling with four outcomes following a peer-to-peer session on a single OMHP: retention in the community, following up on a previous session with a counselor, users’ evaluation of a counselor, and changes in users’ mood. Results show that predictors correlate negatively with community retention but positively with users following up with and giving higher evaluations to individual counselors. We suggest actionable insights for therapy platform design and outcome measurement based on findings that the relationship between predictors and outcomes of successful conversations depends on differences in measurement construct and operationalization.

Recently, large-scale pre-trained Vision-and-Language (VL) foundation models have demonstrated remarkable capabilities in many zero-shot downstream tasks, achieving competitive results for recognizing objects defined by as little as short text prompts. However, it has also been shown that VL models are still brittle in Structured VL Concept (SVLC) reasoning, such as the ability to recognize object attributes, states, and inter-object relations. This leads to reasoning mistakes, which need to be corrected as they occur by teaching VL models the missing SVLC skills; often this must be done using private data where the issue was found, which naturally leads to a data-free continual (no task-id) VL learning setting. In this work, we introduce the first Continual Data-Free Structured VL Concepts Learning (ConStruct-VL) benchmark11Our code is publicly available at https://github.com/jamessealesmith/ConStruct-VL and show it is challenging for many existing data-free CL strategies. We, therefore, propose a data-free method comprised of a new approach of Adversarial Pseudo-Replay (APR) which generates adversarial reminders of past tasks from past task models. To use this method efficiently, we also propose a continual parameter-efficient Layered-LoRA (LaLo) neural architecture allowing no-memory-cost access to all past models at train time. We show this approach outperforms all data-free methods by as much as ~ 7% while even matching some levels of experience-replay (prohibitive for applications where data-privacy must be preserved).

Multilingual transformer-based models demonstrate remarkable zero and few-shot transfer across languages by learning and reusing language-agnostic features.However, as a fixed-size model acquires more languages, its performance across all languages degrades.Those who attribute this interference phenomenon to limited model capacity address the problem by adding additional parameters, despite evidence that transformer-based models are overparameterized.In this work, we show that it is possible to reduce interference by instead identifying and pruning language-specific attention heads.First, we use Shapley Values, a credit allocation metric from coalitional game theory, to identify attention heads that introduce interference.Then, we show that pruning such heads from a fixed model improves performance for a target language on both sentence classification and structural prediction.Finally, we provide insights on language-agnostic and language-specific attention heads using attention visualization. 1

Large language models (LLMs) have achieved significant progress from pre-training on and memorizing a wide range of textual data, however, this process might suffer from privacy issues and violations of data protection regulations.As a result, the ability to easily remove data related to individual users from such models while not deteriorating their predictive quality after the removal becomes increasingly important.To address these issues, in this work, we propose an efficient unlearning framework that could efficiently update LLMs without having to retrain the whole model after data removals, by introducing lightweight unlearning layers learned with a selective teacher-student objective into the transformers.In addition, we introduce a fusion mechanism to effectively combine different unlearning layers that learns to forget different sets of data to handle a sequence of forgetting operations.Experiments on classification and generation tasks demonstrate the effectiveness of our proposed methods compared to the state-of-the-art baselines 1 .

Diffusion models that are based on iterative denoising have been recently proposed and leveraged in various generation tasks like image generation.Whereas, as a way inherently built for continuous data, existing diffusion models still have some limitations in modeling discrete data, e.g., languages.For example, the generally used Gaussian noise can not handle the discrete corruption well, and the objectives in continuous spaces fail to be stable for textual data in the diffusion process especially when the dimension is high.To alleviate these issues, we introduce a novel diffusion model for language modeling, Masked-Diffusion LM, with lower training cost and better performances, inspired by linguistic features in languages.Specifically, we design a linguistic-informed forward process which adds corruptions to the text through strategically soft-masking to better noise the textual data.Also, we directly predict the categorical distribution with cross-entropy loss function in every diffusion step to connect the continuous space and discrete space in a more efficient and straightforward way.Through experiments on 5 controlled generation tasks, we demonstrate that our Masked-Diffusion LM can achieve better generation quality than the state-of-the-art diffusion models with better efficiency.Code is available at https://github. com/SALT-NLP/Masked_Diffusioin_LM.

Recent work has aimed to capture nuances of human behavior by using LLMs to simulate responses from particular demographics in settings like social science experiments and public opinion surveys.However, there are currently no established ways to discuss or evaluate the quality of such LLM simulations.Moreover, there is growing concern that these simulations are flattened caricatures of the personas that they aim to simulate, failing to capture the multidimensionality of people and perpetuating stereotypes.To bridge these gaps, we present CoMPosT, a framework to characterize LLM simulations using four dimensions: Context, Model, Persona, and Topic.We use this framework to measure open-ended LLM simulations' susceptibility to caricature, defined via two criteria: individuation and exaggeration.We evaluate the level of caricature in scenarios from existing work on LLM simulations.We find that for GPT-4, simulations of certain demographics (political and marginalized groups) and topics (general, uncontroversial) are highly susceptible to caricature.

Teachers' growth mindset supportive language (GMSL)-rhetoric emphasizing that one's skills can be improved over time-has been shown to significantly reduce disparities in academic achievement and enhance students' learning outcomes.Although teachers espouse growth mindset principles, most find it difficult to adopt GMSL in their practice due the lack of effective coaching in this area.We explore whether large language models (LLMs) can provide automated, personalized coaching to support teachers' use of GMSL.We establish an effective coaching tool to reframe unsupportive utterances to GMSL by developing (i) a parallel dataset containing GMSL-trained teacher reframings of unsupportive statements with an accompanying annotation guide, (ii) a GMSL prompt framework to revise teachers' unsupportive language, and (iii) an evaluation framework grounded in psychological theory for evaluating GMSL with the help of students and teachers.1 We conduct a large-scale evaluation involving 174 teachers and 1,006 students, finding that both teachers and students perceive GMSL-trained teacher and model reframings as more effective in fostering a growth mindset and promoting challenge-seeking behavior, among other benefits.We also find that model-generated reframings outperform those from the GMSL-trained teachers.These results show promise for harnessing LLMs to provide automated GMSL feedback for teachers and, more broadly, LLMs' potentiality for supporting students' learning in the classroom.Our findings also demonstrate the benefit of largescale human evaluations when applying LLMs in educational domains.

Is aesthetic impact different from beauty?Is visual salience a reflection of its capacity for effective communication?We present Impressions, 1 a novel dataset through which to investigate the semiotics of images, and how specific visual features and design choices can elicit specific emotions, thoughts and beliefs.We posit that the impactfulness of an image extends beyond formal definitions of aesthetics, to its success as a communicative act, where style contributes as much to meaning formation as the subject matter.However, prior image captioning datasets are not designed to empower state-of-the-art architectures to model potential human impressions or interpretations of images.To fill this gap, we design an annotation task heavily inspired by image analysis techniques in the Visual Arts to collect 1,440 image-caption pairs and 4,320 unique annotations exploring impact, pragmatic image description, impressions, and aesthetic design choices.We show that existing multimodal image captioning and conditional generation models struggle to simulate plausible human responses to images.However, this dataset significantly improves their ability to model impressions and aesthetic evaluations of images through fine-tuning and few-shot adaptation.

Annotated data plays a critical role in Natural Language Processing (NLP) in training models and evaluating their performance.Given recent developments in Large Language Models (LLMs), models such as ChatGPT demonstrate zero-shot capability on many text-annotation tasks, comparable with or even exceeding human annotators.Such LLMs can serve as alternatives for manual annotation, due to lower costs and higher scalability.However, limited work has leveraged LLMs as complementary annotators, nor explored how annotation work is best allocated among humans and LLMs to achieve both quality and cost objectives.We propose CoAnnotating, a novel paradigm for Human-LLM co-annotation of unstructured texts at scale.Under this framework, we utilize uncertainty to estimate LLMs' annotation capability.Our empirical study shows CoAnnotating to be an effective means to allocate work from results on different datasets, with up to 21% performance improvement over random baseline.For code implementation, see https: //github.com/SALT-NLP/CoAnnotating.

Existing large language models (LLMs) that mainly focus on Standard American English (SAE) often lead to significantly worse performance when being applied to other English dialects.While existing mitigations tackle discrepancies for individual target dialects, they assume access to high-accuracy dialect identification systems.The boundaries between dialects are inherently flexible, making it difficult to categorize language into discrete predefined categories.In this work, we propose DADA (Dialect Adaptation via Dynamic Aggregation), a modular approach to imbue SAE-trained models with multi-dialectal robustness by composing adapters which handle specific linguistic features.The compositional architecture of DADA allows for both targeted adaptation to specific dialect variants and simultaneous adaptation to various dialects.We show that DADA is effective for both single task and instruction finetuned language models, offering an extensible and interpretable framework for adapting existing LLMs to different English dialects.

Weicheng Ma, Henry Scheible, Brian Wang, Goutham Veeramachaneni, Pratim Chowdhary, Alan Sun, Andrew Koulogeorge, Lili Wang, Diyi Yang, Soroush Vosoughi. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. 2023.

Spurred by advancements in scale, large language models (LLMs) have demonstrated the ability to perform a variety of natural language processing (NLP) tasks zero-shot-i.e., without adaptation on downstream data.Recently, the debut of ChatGPT 1 has drawn a great deal of attention from the natural language processing (NLP) community due to the fact that it can generate high-quality responses to human input and self-correct previous mistakes based on subsequent conversations.However, it is not yet known whether ChatGPT can serve as a generalist model that can perform many NLP tasks zero-shot.In this work, we empirically analyze the zero-shot learning ability of ChatGPT by evaluating it on 20 popular NLP datasets covering 7 representative task categories.With extensive empirical studies, we demonstrate both the effectiveness and limitations of the current version of ChatGPT.We find that ChatGPT performs well on many tasks favoring reasoning capabilities (e.g., arithmetic reasoning) while it still faces challenges when solving specific tasks such as sequence tagging.We additionally provide in-depth analysis through qualitative case studies.

Large Language Models (LLMs) are trained on corpora disproportionally weighted in favor of Standard American English.As a result, speakers of other dialects experience significantly more failures when interacting with these technologies.In practice, these speakers often accommodate their speech to be better understood.Our work shares the belief that language technologies should be designed to accommodate the diversity in English dialects and not the other way around.However, prior works on dialect struggle with generalizing to evolving and emerging dialects in a scalable manner.To fill this gap, our method, Hyper-LoRA, leverages expert linguistic knowledge to enable resource-efficient adaptation via hypernetworks.By disentangling dialect-specific and cross-dialectal information, HyperLoRA improves generalization to unseen dialects in a task-agnostic fashion.Not only is HyperLoRA more scalable in the number of parameters, but it also achieves the best or most competitive performance across 5 dialects in a zero-shot setting.In this way, our approach facilitates access to language technology for billions of English dialect speakers who are traditionally underrepresented.

Developing an educational test can be expensive and time-consuming, as each item must be written by experts and then evaluated by collecting hundreds of student responses.Moreover, many tests require multiple distinct sets of questions administered throughout the school year to closely monitor students' progress, known as parallel tests.In this study, we focus on tests of silent sentence reading efficiency, used to assess students' reading ability over time.To generate high-quality parallel tests, we propose to fine-tune large language models (LLMs) to simulate how previous students would have responded to unseen items.With these simulated responses, we can estimate each item's difficulty and ambiguity.We first use GPT-4 to generate new test items following a list of expert-developed rules and then apply a fine-tuned LLM to filter the items based on criteria from psychological measurements.We also propose an optimal-transport-inspired technique for generating parallel tests and show the generated tests closely correspond to the original test's difficulty and reliability based on crowdworker responses.Our evaluation of a generated test with 234 students from grades 2 to 8 produces test scores highly correlated (r=0.93) to those of a standard test form written by human experts and evaluated across thousands of K-12 students.

At the end of October 2022, Elon Musk concluded his acquisition of Twitter. In the weeks and months before that, several questions were publicly discussed that were not only of interest to the platform's future buyers, but also of high relevance to the Computational Social Science research community. For example, how many active users does the platform have? What percentage of accounts on the site are bots? And, what are the dominating topics and sub-topical spheres on the platform? In a globally coordinated effort of 80 scholars to shed light on these questions, and to offer a dataset that will equip other researchers to do the same, we have collected all 375 million tweets published within a 24-hour time period starting on September 21, 2022. To the best of our knowledge, this is the first complete 24-hour Twitter dataset that is available for the research community. With it, the present work aims to accomplish two goals. First, we seek to answer the aforementioned questions and provide descriptive metrics about Twitter that can serve as references for other researchers. Second, we create a baseline dataset for future research that can be used to study the potential impact of the platform's ownership change.

Social media (i.e., Reddit) users are overloaded with people’s opinions when viewing discourses about divisive topics. Traditional user interfaces in such media present those opinions in a linear structure, which can limit users in viewing diverse social opinions at scale. Prior work has recognized this limitation, that the linear structure can reinforce biases, where a certain point of view becomes widespread simply because many viewers seem to believe it. This limitation can make it difficult for users to have a truly conversational mode of mediated discussion. Thus, when designing a user interface for viewing people’s opinions, we should consider ways to mitigate selective exposure to information and polarization of opinions. We conducted a needs-finding study with 11 Reddit users, who follow climate change threads and make posts and comments regularly. In the study, we aimed to understand key limitations in people viewing online controversial discourses and to extract design implications to address these problems. Our findings discuss potential future directions to address these problems.

As video-sharing social-media platforms have increased in popularity, a 'creator economy' has emerged in which platform users make online content to share with wide audiences, often for profit. As the creator economy has risen in popularity, so have concerns of racism and discrimination on social media. Black content creators across multiple platforms have identified challenges with racism and discrimination, perpetuated by platform users, companies that collaborate with creators for sponsored content, and the algorithms governing these platforms. In this work, we provide a qualitative study of the experiences of Black content creators on one video-sharing platform, TikTok. We conduct 12 semi-structured interviews with Black TikTok content creators to understand their experiences, identify the challenges they face, and understand their perceptions of the platform. We find that some common challenges include: content moderation, monetization, harassment and bullying from viewers, lack of transparency of recommendation and filtering algorithms, and the perception that content from Black creators is treated unfairly by those algorithms. We then suggest design interventions to mitigate the challenges, bolster positive aspects, and overall cultivate an inclusive algorithmic experience for Black creators on TikTok

Abstract NLP has achieved great progress in the past decade through the use of neural models and large labeled datasets. The dependence on abundant data prevents NLP models from being applied to low-resource settings or novel tasks where significant time, money, or expertise is required to label massive amounts of textual data. Recently, data augmentation methods have been explored as a means of improving data efficiency in NLP. To date, there has been no systematic empirical overview of data augmentation for NLP in the limited labeled data setting, making it difficult to understand which methods work in which settings. In this paper, we provide an empirical survey of recent progress on data augmentation for NLP in the limited labeled data setting, summarizing the landscape of methods (including token-level augmentations, sentence-level augmentations, adversarial augmentations, and hidden-space augmentations) and carrying out experiments on 11 datasets covering topics/news classification, inference tasks, paraphrasing tasks, and single-sentence tasks. Based on the results, we draw several conclusions to help practitioners choose appropriate augmentations in different settings and discuss the current challenges and future directions for limited data learning in NLP.

The study of network robustness is a critical tool in the characterization and sense making of complex interconnected systems such as infrastructure, communication and social networks. While significant research has been conducted in these areas, gaps in the surveying literature still exist. Answers to key questions are currently scattered across multiple scientific fields and numerous papers. In this survey, we distill key findings across numerous domains and provide researchers crucial access to important information by(1) summarizing and comparing recent and classical graph robustness measures; (2) exploring which robustness measures are most applicable to different categories of networks (e.g., social, infrastructure); (3) reviewing common network attack strategies, and summarizing which attacks are most effective across different network topologies; and (4) extensive discussion on selecting defense techniques to mitigate attacks across a variety of networks. This survey guides researchers and practitioners in navigating the expansive field of network robustness, while summarizing answers to key questions. We conclude by highlighting current research directions and open problems.

Keeping track of scientific challenges, advances and emerging directions is a fundamental part of research. However, researchers face a flood of papers that hinders discovery of important knowledge. In biomedicine, this directly impacts human lives. To address this problem, we present a novel task of extraction and search of scientific challenges and directions, to facilitate rapid knowledge discovery. We construct and release an expert-annotated corpus of texts sampled from full-length papers, labeled with novel semantic categories that generalize across many types of challenges and directions. We focus on a large corpus of interdisciplinary work relating to the COVID-19 pandemic, ranging from biomedicine to areas such as AI and economics. We apply a model trained on our data to identify challenges and directions across the corpus and build a dedicated search engine. In experiments with 19 researchers and clinicians using our system, we outperform a popular scientific search engine in assisting knowledge discovery. Finally, we show that models trained on our resource generalize to the wider biomedical domain and to AI papers, highlighting its broad utility. We make our data, model and search engine publicly available.

Ying Xu, Dakuo Wang, Mo Yu, Daniel Ritchie, Bingsheng Yao, Tongshuang Wu, Zheng Zhang, Toby Li, Nora Bradford, Branda Sun, Tran Hoang, Yisi Sang, Yufang Hou, Xiaojuan Ma, Diyi Yang, Nanyun Peng, Zhou Yu, Mark Warschauer. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2022.

Continual learning is essential for real-world deployment when there is a need to quickly adapt the model to new tasks without forgetting knowledge of old tasks.Existing work on continual sequence generation either always reuses existing parameters to learn new tasks, which is vulnerable to catastrophic forgetting on dissimilar tasks, or blindly adds new parameters for every new task, which could prevent knowledge sharing between similar tasks.To get the best of both worlds, in this work, we propose continual sequence generation with adaptive compositional modules to adaptively add modules in transformer architectures and compose both old and new modules for new tasks.We also incorporate pseudo experience replay to facilitate knowledge transfer in those shared modules.Experiment results on various sequences of generation tasks show that our framework can adaptively add modules or reuse modules based on task similarity, outperforming state-of-the-art baselines in terms of both performance and parameter efficiency.We make our code public at https://github.com/GT-SALT/ Adaptive-Compositional-Modules.

English Natural Language Understanding (NLU) systems have achieved great performances and even outperformed humans on benchmarks like GLUE and SuperGLUE.However, these benchmarks contain only textbook Standard American English (SAE).Other dialects have been largely overlooked in the NLP community.This leads to biased and inequitable NLU systems that serve only a sub-population of speakers.To understand disparities in current models and to facilitate more dialect-competent NLU systems, we introduce the VernAcular Language Understanding Evaluation (VALUE) benchmark, a challenging variant of GLUE that we created with a set of lexical and morphosyntactic transformation rules.In this initial release (V.1), we construct rules for 11 features of African American Vernacular English (AAVE), and we recruit fluent AAVE speakers to validate each feature transformation via linguistic acceptability judgments in a participatory design manner.Experiments show that these new dialectal features can lead to a drop in model performance.

Sentiment transfer is one popular example of a text style transfer task, where the goal is to reverse the sentiment polarity of a text.With a sentiment reversal comes also a reversal in meaning.We introduce a different but related task called positive reframing in which we neutralize a negative point of view and generate a more positive perspective for the author without contradicting the original meaning.Our insistence on meaning preservation makes positive reframing a challenging and semantically rich task.To facilitate rapid progress, we introduce a large-scale benchmark, POSITIVE PSY-CHOLOGY FRAMES, with 8,349 sentence pairs and 12,755 structured annotations to explain positive reframing in terms of six theoreticallymotivated reframing strategies.Then we evaluate a set of state-of-the-art text style transfer models, and conclude by discussing key challenges and directions for future work.To download the data, see https://github. com/GT-SALT/positive-frames

Content Warning: some examples in this paper may be offensive or upsetting.Conversational agents have come increasingly closer to human competence in open-domain dialogue settings; however, such models can reflect insensitive, hurtful, or entirely incoherent viewpoints that erode a user's trust in the moral integrity of the system.Moral deviations are difficult to mitigate because moral judgments are not universal, and there may be multiple competing judgments that apply to a situation simultaneously.In this work, we introduce a new resource, not to authoritatively resolve moral ambiguities, but instead to facilitate systematic understanding of the intuitions, values and moral judgments reflected in the utterances of dialogue systems.The MORAL INTEGRITY CORPUS, MIC , is such a resource, which captures the moral assumptions of 38k prompt-reply pairs, using 99k distinct Rules of Thumb (RoTs).Each RoT reflects a particular moral conviction that can explain why a chatbot's reply may appear acceptable or problematic.We further organize RoTs with a set of 9 moral and social attributes and benchmark performance for attribute classification.Most importantly, we show that current neural language models can automatically generate new RoTs that reasonably describe previously unseen interactions, but they still struggle with certain scenarios.Our findings suggest that MIC will be a useful resource for understanding and language models' implicit moral assumptions and flexibly benchmarking the integrity of conversational agents.

Large language models are increasingly mediating, modifying, and even generating messages for users, but the receivers of these messages may not be aware of the involvement of AI. To examine this emerging direction of AI-Mediated Communication (AI-MC), we investigate people’s perceptions of AI written messages. We analyze how such perceptions change in accordance with the interpersonal emphasis of a given message. We conducted both large-scale surveys and in-depth interviews to investigate how a diverse set of factors influence people’s perceived trust in AI-mediated writing of emails. We found that people’s trust in email writers decreased when they were told that AI was involved in the writing process. Surprisingly trust increased when AI was used for writing more interpersonal emails (as opposed to more transactional ones). Our study provides insights regarding how people perceive AI-MC and has practical design implications on building AI-based products to aid human interlocutors in communication1.

People write personalized greeting cards on various occasions. While prior work has studied gender roles in greeting card messages, systematic analysis at scale and tools for raising the awareness of gender stereotyping remain under-investigated. To this end, we collect a large greeting card message corpus covering three different occasions (birthday, Valentine’s Day and wedding) from three sources (exemplars from greeting message websites, real-life greetings from social media and language model generated ones). We uncover a wide range of gender stereotypes in this corpus via topic modeling, odds ratio and Word Embedding Association Test (WEAT). We further conduct a survey to understand people’s perception of gender roles in messages from this corpus and if gender stereotyping is a concern. The results show that people want to be aware of gender roles in the messages, but remain unconcerned unless the perceived gender roles conflict with the recipient’s true personality. In response, we developed GreetA, an interactive visualization and writing assistant tool to visualize fine-grained topics in greeting card messages drafted by the users and the associated gender perception scores, but without suggesting text changes as an intervention.

This paper proposes a simple yet effective interpolation-based data augmentation approach termed DoubleMix, to improve the robustness of models in text classification. DoubleMix first leverages a couple of simple augmentation operations to generate several perturbed samples for each training data, and then uses the perturbed data and original data to carry out a two-step interpolation in the hidden space of neural models. Concretely, it first mixes up the perturbed data to a synthetic sample and then mixes up the original data and the synthetic perturbed data. DoubleMix enhances models’ robustness by learning the “shifted” features in hidden space. On six text classification benchmark datasets, our approach outperforms several popular text augmentation methods including token-level, sentence-level, and hidden-level data augmentation techniques. Also, experiments in low-resource settings show our approach consistently improves models’ performance when the training data is scarce. Extensive ablation studies and case studies confirm that each component of our approach contributes to the final performance and show that our approach exhibits superior performance on challenging counterexamples. Additionally, visual analysis shows that text features generated by our approach are highly interpretable.

Demonstration-based learning has shown great potential in stimulating pretrained language models' ability under limited data scenario.Simply augmenting the input with some demonstrations can significantly improve performance on few-shot NER.However, why such demonstrations are beneficial remains unclear since there is no explicit alignment between the demonstrations and the predictions.In this paper, we design pathological demonstrations by gradually removing intuitively useful information from the standard ones to take a deep dive of the robustness of demonstration-based sequence labeling and show that (1) demonstrations composed of random tokens still make the model a better few-shot learner; (2) the length of random demonstrations and the relevance of random tokens are the main factors affecting the performance; (3) demonstrations increase the confidence of model predictions on captured superficial patterns.We have publicly released our code at https://github.com/ SALT-NLP/RobustDemo.

In a fair world, people have equitable opportunities to education, to conduct scientific research, to publish, and to get credit for their work, regardless of where they live.However, it is common knowledge among researchers that a vast number of papers accepted at top NLP venues come from a handful of western countries and (lately) China; whereas, very few papers from Africa and South America get published.Similar disparities are also believed to exist for paper citation counts.In the spirit of "what we do not measure, we cannot improve", this work asks a series of questions on the relationship between geographical location and publication success (acceptance in top NLP venues and citation impact).We first created a dataset of 70,000 papers from the ACL Anthology, extracted their meta-information, and generated their citation network.We then show that not only are there substantial geographical disparities in paper acceptance and citation but also that these disparities persist even when controlling for a number of variables such as venue of publication and sub-field of NLP.Further, despite some steps taken by the NLP community to improve geographical diversity, we show that the disparity in publication metrics across locations is still on an increasing trend since the early 2000s.We release our code and dataset here

Raj Shah, Kunal Chawla, Dheeraj Eidnani, Agam Shah, Wendi Du, Sudheer Chava, Natraj Raman, Charese Smiley, Jiaao Chen, Diyi Yang. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing. 2022.

Document authoring involves a lengthy revision process, marked by individual edits that are frequently linked to comments. Modeling the relationship between edits and comments leads to a better understanding of document evolution, potentially benefiting applications such as content summarization, and task triaging. Prior work on understanding revisions has primarily focused on classifying edit intents, but falling short of a deeper understanding of the nature of these edits. In this paper, we present explore the challenge of describing an edit at two levels: identifying the edit intent, and describing the edit using free-form text. We begin by defining a taxonomy of general edit intents and introduce a new dataset of full revision histories of Wikipedia pages, annotated with each revision’s edit intent. Using this dataset, we train a classifier that achieves a 90% accuracy in identifying edit intent. We use this classifier to train a distantly-supervised model that generates a high-level description of a revision in free-form text. Our experimental results show that incorporating edit intent information aids in generating better edit descriptions. We establish a set of baselines for the edit description task, achieving a best score of 28 ROUGE, thus demonstrating the effectiveness of our layered approach to edit understanding.

We propose a novel knowledge grounded dialogue (interview) dataset SPORTSINTERVIEW set in the domain of sports interview. Our dataset contains two types of external knowledge sources as knowledge grounding, and is rich in content, containing about 150K interview sessions and 34K distinct interviewees. Compared to existing knowledge grounded dialogue datasets, our interview dataset is larger in size, comprises natural dialogues revolving around real-world sports matches, and have more than one dimension of external knowledge linking. We performed several experiments on SPORTSINTERVIEW and found that models such as BART fine-tuned on our dataset are able to learn lots of relevant domain knowledge and generate meaningful sentences (questions or responses). However, their performance is still far from humans (by comparing to gold sentences in the dataset) and hence encourages future research utilizing SPORTSINTERVIEW.

As NLP models achieved state-of-the-art performances over benchmarks and gained wide applications, it has been increasingly important to ensure the safe deployment of these models in the real world, e.g., making sure the models are robust against unseen or challenging scenarios.Despite robustness being an increasingly studied topic, it has been separately explored in applications like vision and NLP, with various definitions, evaluation and mitigation strategies in multiple lines of research.In this paper, we aim to provide a unifying survey of how to define, measure and improve robustness in NLP.We first connect multiple definitions of robustness, then unify various lines of work on identifying robustness failures and evaluating models' robustness.Correspondingly, we present mitigation strategies that are data-driven, model-driven, and inductive-prior-based, with a more systematic view of how to effectively improve robustness in NLP models.Finally, we conclude by outlining open challenges and future directions to motivate further research in this area.

Warning: This paper contains content that is offensive and may be upsetting.Biased or toxic speech can be harmful to various demographic groups.Therefore, it is not only important for models to detect these speech, but to also output explanations of why a given text is toxic.Previous literature has mostly focused on classifying and detecting toxic speech, and existing efforts on explaining stereotypes in toxic speech mainly use standard text generation approaches, resulting in generic and repetitive explanations.Building on these prior works, we introduce a novel knowledgeinformed encoder-decoder framework to utilize multiple knowledge sources to generate implications of biased text.Experiments show that our knowledge informed models outperform prior state-of-the-art models significantly, and can generate detailed explanations of stereotypes in toxic speech compared to baselines, both quantitatively and qualitatively.

Although sequence-to-sequence models often achieve good performance in semantic parsing for i.i.d.data, their performance is still inferior in compositional generalization.Several data augmentation methods have been proposed to alleviate this problem.However, prior work only leveraged superficial grammar or rules for data augmentation, which resulted in limited improvement.We propose to use subtree substitution for compositional data augmentation, where we consider subtrees with similar semantic functions as exchangeable.Our experiments showed that such augmented data led to significantly better performance on SCAN and GEOQUERY, and reached new SOTA on compositional split of GEOQUERY.

Data augmentation is an effective approach to tackle over-fitting.Many previous works have proposed different data augmentations strategies for NLP, such as noise injection, word replacement, back-translation etc.Though effective, they missed one important characteristic of language-compositionality, meaning of a complex expression is built from its sub-parts.Motivated by this, we propose a compositional data augmentation approach for natural language understanding called TreeMix.Specifically, TreeMix leverages constituency parsing tree to decompose sentences into constituent sub-structures and the Mixup data augmentation technique to recombine them to generate new sentences.Compared with previous approaches, TreeMix introduces greater diversity to the samples generated and encourages models to learn compositionality of NLP data.Extensive experiments on text classification and SCAN demonstrate that TreeMix outperforms current state-of-the-art data augmentation methods.𝑹𝑹𝑹𝑹𝑹𝑹𝑹𝑹𝑹𝑹𝑹𝑹 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝑹𝑹 𝒔𝒔𝑹𝑹𝑹𝑹𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔 { x 𝑖𝑖 : " They will find little interest in this poor film.", y

Millions of people participate in online peer-to-peer support sessions, yet there has been little prior research on systematic psychology-based evaluations of fine-grained peer-counselor behavior in relation to client satisfaction. This paper seeks to bridge this gap by mapping peer-counselor chat-messages to motivational interviewing (MI) techniques. We annotate 14,797 utterances from 734 chat conversations using 17 MI techniques and introduce four new interviewing codes such as ''chit-chat'' and ''inappropriate'' to account for the unique conversational patterns observed on online platforms. We automate the process of labeling peer-counselor responses to MI techniques by fine-tuning large domain-specific language models and then use these automated measures to investigate the behavior of the peer counselors via correlational studies. Specifically, we study the impact of MI techniques on the conversation ratings to investigate the techniques that predict clients' satisfaction with their counseling sessions. When counselors use techniques such as reflection and affirmation, clients are more satisfied. Examining volunteer counselors' change in usage of techniques suggest that counselors learn to use more introduction and open questions as they gain experience. This work provides a deeper understanding of the use of motivational interviewing techniques on peer-to-peer counselor platforms and sheds light on how to build better training programs for volunteer counselors on online platforms.

Since the COVID-19 pandemic, businesses have faced unprecedented challenges when trying to remain open. Because COVID-19 spreads through aerosolized droplets, businesses were forced to distance their services; in some cases, distancing may have involved moving business services online. In this work, we explore digitization strategies used by small businesses that remained open during the pandemic, and survey/interview small businesses owners to understand preliminary challenges associated with moving online. Furthermore, we analyze payments from 400K businesses across Japan, Australia, United States, Great Britain, and Canada. Following initial government interventions, we observe (at minimum for each country) a 47% increase in digitizing businesses compared to pre-pandemic levels, with about 80% of surveyed businesses digitizing in under a week. From both our quantitative models and our surveys/interviews, we find that businesses rapidly digitized at the start of the pandemic in preparation of future uncertainty. We also conduct a case-study of initial digitization in the United States, examining finer relationships between specific government interventions, business sectors, political orientation, and resulting digitization shifts. Finally, we discuss the implications of rapid & widespread digitization for small businesses in the context of usability challenges and interpersonal interactions, while highlighting potential shifts in pre-existing social norms.

Abstract A fundamental goal of scientific research is to learn about causal relationships. However, despite its critical role in the life and social sciences, causality has not had the same importance in Natural Language Processing (NLP), which has traditionally placed more emphasis on predictive tasks. This distinction is beginning to fade, with an emerging area of interdisciplinary research at the convergence of causal inference and language processing. Still, research on causality in NLP remains scattered across domains without unified definitions, benchmark datasets and clear articulations of the challenges and opportunities in the application of causal inference to the textual domain, with its unique properties. In this survey, we consolidate research across academic areas and situate it in the broader NLP landscape. We introduce the statistical challenge of estimating causal effects with text, encompassing settings where text is used as an outcome, treatment, or to address confounding. In addition, we explore potential uses of causal inference to improve the robustness, fairness, and interpretability of NLP models. We thus provide a unified overview of causal inference for the NLP community.1

Modeling persuasive language has the potential to better facilitate our decision-making processes. Despite its importance, computational modeling of persuasion is still in its infancy, largely due to the lack of benchmark datasets that can provide quantitative labels of persuasive strategies to expedite this line of research. To this end, we introduce a large-scale multi-domain text corpus for modeling persuasive strategies in good-faith text requests. Moreover, we design a hierarchical weakly-supervised latent variable model that can leverage partially labeled data to predict such associated persuasive strategies for each sentence, where the supervision comes from both the overall document-level labels and very limited sentence-level labels. Experimental results showed that our proposed method outperformed existing semi-supervised baselines significantly. We have publicly released our code at https://github.com/GT-SALT/Persuasion_Strategy_WVAE.

Jiaao Chen, Dinghan Shen, Weizhu Chen, Diyi Yang. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.

The spread of COVID-19 has sparked racism and hate on social media targeted towards Asian communities. However, little is known about how racial hate spreads during a pandemic and the role of counterspeech in mitigating this spread. In this work, we study the evolution and spread of anti-Asian hate speech through the lens of Twitter. We create COVID-HATE, the largest dataset of anti-Asian hate and counterspeech spanning 14 months, containing over 206 million tweets, and a social network with over 127 million nodes. By creating a novel hand-labeled dataset of 3,355 tweets, we train a text classifier to identify hateful and counterspeech tweets that achieves an average macro-F1 score of 0.832. Using this dataset, we conduct longitudinal analysis of tweets and users. Analysis of the social network reveals that hateful and counterspeech users interact and engage extensively with one another, instead of living in isolated polarized communities. We find that nodes were highly likely to become hateful after being exposed to hateful content in the year 2020. Notably, counterspeech messages discourage users from turning hateful, potentially suggesting a solution to curb hate on web and social media platforms. Data and code is available at http://claws.cc.gatech.edu/covid.

New parents, defined as parents of children between the infant and preschooler stages, are increasingly turning to online media to exchange support and information to help with their life-changing transition. Understanding parents’ discussion online is crucial to the design and development of technologies that can better support their media interaction. This work studies how new parents use online media using a large-scale parenting corpus. To do so, we first employed a card-sorting methodology to identify a set of parenting topics, with which we trained BERT classifiers to automatically identify the topics of Reddit posts. We then investigate at scale what parenting topics were talked about most by new parents, how topics change over the course of their participation, and how interactions with different topics affect members’ engagement in the community. We conclude with implications of our research in designing future research and online parenting communities.

Online mental health communities (OMHCs) are prominent resources for improving people’s mental wellbeing. An immediate cue of such improvement is support-seekers’ satisfaction expressed in their replies to the received comments. However, the comments that seekers find satisfying may change with their community knowledge, e.g., measured by tenure and posting experience in that community. In this paper, we first model the amount of satisfaction conveyed in the support-seekers’ replies to the received comments. Then we quantitatively examine how seekers’ expressed satisfaction is affected by their community knowledge, sought and received support in an OMHC. Results show that support-seekers with more posting experience generally display less contentment to the received comments. Compared to newcomers, higher tenured members express less satisfaction when receiving informational support. We also found that support matching positively predicts seekers’ satisfaction regardless of their community knowledge. Our findings have implications for OMHCs to satisfy support-seekers through their community knowledge.

Network robustness plays a crucial role in our understanding of complex interconnected systems such as transportation, communication, and computer networks. While significant research has been conducted in the area of network robustness, no comprehensive open-source toolbox currently exists to assist researchers and practitioners in this important topic. This lack of available tools hinders reproducibility and examination of existing work, development of new research, and dissemination of new ideas. We contribute TIGER, an open-sourced Python toolbox to address these challenges. TIGER contains 22 graph robustness measures with both original and fast approximate versions; 17 failure and attack strategies; 15 heuristic and optimization-based defense techniques; and 4 simulation tools. By democratizing the tools required to study network robustness, our goal is to assist researchers and practitioners in analyzing their own networks; and facilitate the development of new research in the field. TIGER has been integrated into the Nvidia Data Science Teaching Kit available to educators across the world; and Georgia Tech's Data and Visual Analytics class with over 1,000 students. TIGER is open sourced at: https://github.com/safreita1/TIGER

Abstractive conversation summarization has received growing attention while most current state-of-the-art summarization models heavily rely on human-annotated summaries.To reduce the dependence on labeled summaries, in this work, we present a simple yet effective set of Conversational Data Augmentation (CODA) methods for semisupervised abstractive conversation summarization, such as random swapping/deletion to perturb the discourse relations inside conversations, dialogue-acts-guided insertion to interrupt the development of conversations, and conditional-generation-based substitution to substitute utterances with their paraphrases generated based on the conversation context.To further utilize unlabeled conversations, we combine CODA with two-stage noisy selftraining where we first pre-train the summarization model on unlabeled conversations with pseudo summaries and then fine-tune it on labeled conversations.Experiments conducted on the recent conversation summarization datasets demonstrate the effectiveness of our methods over several state-of-the-art data augmentation baselines.

Mai ElSherief, Caleb Ziems, David Muchlinski, Vaishnavi Anupindi, Jordyn Seybolt, Munmun De Choudhury, Diyi Yang. Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing. 2021.

Interpolation-based regularisation methods for data augmentation have proven to be effective for various tasks and modalities.These methods involve performing mathematical operations over the raw input samples or their latent states representations -vectors that often possess complex hierarchical geometries.However, these operations are performed in the Euclidean space, simplifying these representations, which may lead to distorted and noisy interpolations.We propose HypMix, a novel model-, data-, and modality-agnostic interpolative data augmentation technique operating in the hyperbolic space, which captures the complex geometry of input and hidden state hierarchies better than its contemporaries.We evaluate HypMix on benchmark and low resource datasets across speech, text, and vision modalities, showing that HypMix consistently outperforms state-of-the-art data augmentation techniques.In addition, we demonstrate the use of HypMix in semi-supervised settings.We further probe into the adversarial robustness and qualitative inferences we draw from HypMix that elucidate the efficacy of the Riemannian hyperbolic manifolds for interpolation-based data augmentation.

The availability of corpora has led to significant advances in training semantic parsers in English.Unfortunately, for languages other than English, annotated data is limited and so is the performance of the developed parsers.Recently, pretrained multilingual models have been proven useful for zero-shot cross-lingual transfer in many NLP tasks.What else does it require to apply a parser trained in English to other languages for zero-shot cross-lingual semantic parsing?Will simple language-independent features help?To this end, we experiment with six Discourse Representation Structure (DRS) semantic parsers in English, and generalize them to Italian, German and Dutch, where there are only a small number of manually annotated parses available.Extensive experiments show that despite its simplicity, adding Universal Dependency (UD) relations and Universal POS tags (UPOS) as model-agnostic features achieves surprisingly strong improvement on all parsers.We have publicly released our code at https://github.com/GT-SALT/

Abstractive conversation summarization has received much attention recently.However, these generated summaries often suffer from insufficient, redundant, or incorrect content, largely due to the unstructured and complex characteristics of human-human interactions.To this end, we propose to explicitly model the rich structures in conversations for more precise and accurate conversation summarization, by first incorporating discourse relations between utterances and action triples ("WHO-DOING-WHAT") in utterances through structured graphs to better encode conversations, and then designing a multi-granularity decoder to generate summaries by combining all levels of information.Experiments show that our proposed models outperform state-of-theart methods and generalize well in other domains in terms of both automatic evaluations and human judgments.We have publicly released our code at https://github.com/ GT-SALT/Structure-Aware-BART.

Natural language processing (NLP) applications are now more powerful and ubiquitous than ever before.With rapidly developing (neural) models and ever-more available data, current NLP models have access to more information than any human speaker during their life.Still, it would be hard to argue that NLP models have reached human-level capacity.In this position paper, we argue that the reason for the current limitations is a focus on information content while ignoring language's social factors.We show that current NLP systems systematically break down when faced with interpreting the social factors of language.This limits applications to a subset of information-related tasks and prevents NLP from reaching human-level performance.At the same time, systems that incorporate even a minimum of social factors already show remarkable improvements.We formalize a taxonomy of seven social factors based on linguistic theory and exemplify current failures and emerging successes for each of them.We suggest that the NLP community address social factors to get closer to the goal of humanlike language understanding.

Yufan Huang, Yanzhe Zhang, Jiaao Chen, Xuezhi Wang, Diyi Yang. Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2021.

Despite the impressive successes of generation and dialogue systems, how to endow a text generation system with particular personality traits to deliver more personalized responses remains under-investigated.In this work, we look at how to generate personalized responses for questions on Reddit by utilizing personalized user profiles and posting histories.Specifically, we release an open-domain single-turn dialog dataset made up of 1.5M conversation pairs together with 300k profiles of users and related comments.We then propose a memory network to generate personalized responses in dialogue that utilizes a novel mechanism of splitting memories: one for user profile meta attributes and the other for user-generated information like comment histories.Experimental results show the quantitative and qualitative improvements of our simple split memory network model over the state-of-the-art response generation baselines.The dataset and code are available here.

Deplatforming refers to the permanent ban of controversial public figures with large followings on social media sites. In recent years, platforms like Facebook, Twitter and YouTube have deplatformed many influencers to curb the spread of offensive speech. We present a case study of three high-profile influencers who were deplatformed on Twitter---Alex Jones, Milo Yiannopoulos, and Owen Benjamin. Working with over 49M tweets, we found that deplatforming significantly reduced the number of conversations about all three individuals on Twitter. Further, analyzing the Twitter-wide activity of these influencers' supporters, we show that the overall activity and toxicity levels of supporters declined after deplatforming. We contribute a methodological framework to systematically examine the effectiveness of moderation interventions and discuss broader implications of using deplatforming as a moderation strategy.

With the widespread use of toxic language online, platforms are increasingly using automated systems that leverage advances in natural language processing to automatically flag and remove toxic comments. However, most automated systems, when detecting and moderating toxic language, do not provide feedback to their users, let alone provide an avenue of recourse for these users to make actionable changes. We present our work, RECAST, an interactive, open-sourced web tool for visualizing these models' toxic predictions, while providing alternative suggestions for flagged toxic language. Our work also provides users with a new path of recourse when using these automated moderation tools. RECAST highlights text responsible for classifying toxicity, and allows users to interactively substitute potentially toxic phrases with neutral alternatives. We examined the effect of RECAST via two large-scale user evaluations, and found that RECAST was highly effective at helping users reduce toxicity as detected through the model. Users also gained a stronger understanding of the underlying toxicity criterion used by black-box models, enabling transparency and recourse. In addition, we found that when users focus on optimizing language for these models instead of their own judgement (which is the implied incentive and goal of deploying automated models), these models cease to be effective classifiers of toxicity compared to human annotations. This opens a discussion for how toxicity detection models work and should work, and their effect on the future of online discourse.

Online health communities (OHCs) have become important resources from which members can obtain social support. Since most benefits of OHCs are provided by its members, it is crucial that OHCs maintain a critical mass of active members. This paper examines temporal changes in members' participation in a cancer-orientedOHC, focusing on the changes in members' motivations and behavior as they transition from newcomers to other roles or when they ultimately leave the community. Our work used mixed methods, combining behavioral log analysis, automated content analysis, surveys and interviews. We found that shifts in members' motivations seemed to be driven by two sources: the internal dynamics common to becoming a member of most online communities and the external needs associated with their cancer journey. When members' disease-driven needs for support decreased, most members quit the site. The motivations of those who stayed shifted from receiving support to providing it to others in the community. As in many online communities, old-timers contributed the vast majority of content. However, they encountered challenges that threatened their commitment, including negative emotions related to other members' deaths, which led them to take leaves of absence from the community or to drop out permanently. Implications for the motivation changes ofOHC members are discussed.

Texts like news, encyclopedias, and some social media strive for objectivity. Yet bias in the form of inappropriate subjectivity — introducing attitudes via framing, presupposing truth, and casting doubt — remains ubiquitous. This kind of bias erodes our collective trust and fuels social conflict. To address this issue, we introduce a novel testbed for natural language generation: automatically bringing inappropriately subjective text into a neutral point of view (“neutralizing” biased text). We also offer the first parallel corpus of biased language. The corpus contains 180,000 sentence pairs and originates from Wikipedia edits that removed various framings, presuppositions, and attitudes from biased sentences. Last, we propose two strong encoder-decoder baselines for the task. A straightforward yet opaque concurrent system uses a BERT encoder to identify subjective words as part of the generation process. An interpretable and controllable modular algorithm separates these steps, using (1) a BERT-based classifier to identify problematic words and (2) a novel join embedding through which the classifier can edit the hidden states of the encoder. Large-scale human evaluation across four domains (encyclopedias, news headlines, books, and political speeches) suggests that these algorithms are a first step towards the automatic identification and reduction of bias.

This paper presents MixText, a semisupervised learning method for text classification, which uses our newly designed data augmentation method called TMix.TMix creates a large amount of augmented training samples by interpolating text in hidden space.Moreover, we leverage recent advances in data augmentation to guess low-entropy labels for unlabeled data, hence making them as easy to use as labeled data.By mixing labeled, unlabeled and augmented data, MixText significantly outperformed current pre-trained and fined-tuned models and other state-ofthe-art semi-supervised learning methods on several text classification benchmarks.The improvement is especially prominent when supervision is extremely limited.

In the course of every member's integration into an online community, a decision must be made to participate for the first time. The challenges of effective recruitment, management, and retention of new users have been extensively explored in social computing research. However, little work has looked at in-the-moment factors that lead users to decide to participate instead of "lurk", conditions which can be shaped to draw new users in at crucial moments. In this work we analyze 183 million messages scraped from chatrooms on the livestreaming platform Twitch in order to understand differences between first-time participants' and regulars' behaviors and to identify conditions that encourage first-time participation. We find that presence of diverse types of users increases likelihood of new participation, with effects depending on the size of the community. We also find that information-seeking behaviors in first-time participation are negatively associated with retention in the short and medium term.

Named Entity Recognition (NER) is one of the first stages in deep language understanding yet current NER models heavily rely on humanannotated data.In this work, to alleviate the dependence on labeled data, we propose a Local Additivity based Data Augmentation (LADA) method for semi-supervised NER, in which we create virtual samples by interpolating sequences close to each other.Our approach has two variations: Intra-LADA and Inter-LADA, where Intra-LADA performs interpolations among tokens within one sentence, and Inter-LADA samples different sentences to interpolate.Through linear additions between sampled training data, LADA creates an infinite amount of labeled data and improves both entity and context learning.We further extend LADA to the semi-supervised setting by designing a novel consistency loss for unlabeled data.Experiments conducted on two NER benchmarks demonstrate the effectiveness of our methods over several strong baselines.We have publicly released our code at

Text summarization is one of the most challenging and interesting problems in NLP.Although much attention has been paid to summarizing structured text like news reports or encyclopedia articles, summarizing conversations-an essential part of humanhuman/machine interaction where most important pieces of information are scattered across various utterances of different speakersremains relatively under-investigated.This work proposes a multi-view sequence-tosequence model by first extracting conversational structures of unstructured daily chats from different views to represent conversations and then utilizing a multi-view decoder to incorporate different views to generate dialogue summaries.Experiments on a large-scale dialogue summarization corpus demonstrated that our methods significantly outperformed previous state-of-the-art models via both automatic evaluations and human judgment.We also discussed specific challenges that current approaches faced with this task.We have publicly released our code at https://github.com/GT-SALT/ Multi-View-Seq2Seq. ConversationTopic View Stage View James: Hey!I have

Ankur Parikh, Xuezhi Wang, Sebastian Gehrmann, Manaal Faruqui, Bhuwan Dhingra, Diyi Yang, Dipanjan Das. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2020.

Existing approaches to disfluency detection heavily depend on human-annotated data.Numbers of data augmentation methods have been proposed to alleviate the dependence on labeled data.However, current augmentation approaches such as random insertion or repetition fail to resemble training corpus well and usually resulted in unnatural and limited types of disfluencies.In this work, we propose a simple Planner-Generator based disfluency generation model to generate natural and diverse disfluent texts as augmented data, where the Planner decides on where to insert disfluent segments and the Generator follows the prediction to generate corresponding disfluent segments.We further utilize this augmented data for pretraining and leverage it for the task of disfluency detection.Experiments demonstrated that our two-stage disfluency generation model outperforms existing baselines; those disfluent sentences generated significantly aided the task of disfluency detection and led to state-of-the-art performance on Switchboard corpus.We have publicly released our code at https://github.com/GT-SALT/ Disfluency-Generation-and-Detection.

Attracting and retaining newcomers is critical and challenging for online production communities such as Wikipedia, both because volunteers need specialized training and are likely to leave before being integrated into the community. In response to these challenges, the Wikimedia Foundation started the Wiki Education Project (Wiki Ed), an online program in which college students edit Wikipedia articles as class assignments. The Wiki Ed program incorporates many components of institutional socialization, a process many conventional organizations successfully use to integrate new employees through formalized on-boarding practices. Research has not adequately investigated whether Wiki Ed and similar programs are effective ways to integrate volunteers in online communities, and, if so, the mechanisms involved. This paper evaluates the Wiki Ed program by comparing 16,819 student editors in 770 Wiki Ed classes with new editors who joined Wikipedia in the conventional way. The evaluation shows that the Wiki Ed students did more work, improved articles more, and were more committed to Wikipedia. For example, compared to new editors who joined Wikipedia in the conventional way they were twice as likely to still be editing Wikipedia a year after their Wiki Ed class was finished. Further, students in classrooms that encouraged joint activity, a key component of institutional socialization, produced better quality work than those in classrooms where students worked independently. These findings are consistent with an interpretation that the Wiki Ed program was successful because it incorporated elements of institutionalized socialization.

Participants in online communities often enact different roles when participating in their communities. For example, some in cancer support communities specialize in providing disease-related information or socializing new members. This work clusters the behavioral patterns of users of a cancer support community into specific functional roles. Based on a series of quantitative and qualitative evaluations, this research identified eleven roles that members occupy, such as welcomer and story sharer. We investigated role dynamics, including how roles change over members' lifecycles, and how roles predict long-term participation in the community. We found that members frequently change roles over their history, from ones that seek resources to ones offering help, while the distribution of roles is stable over the community's history. Adopting certain roles early on predicts members' continued participation in the community. Our methodology will be useful for facilitating better use of members' skills and interests in support of community-building efforts.

People with health concerns go to online health support groups to obtain help and advice. To do so, they frequently disclose personal details, many times in public. Although research in non-health settings suggests that people self-disclose less in public than in private, this pattern may not apply to health support groups where people want to get relevant help. Our work examines how the use of private and public channels influences members' self-disclosure in an online cancer support group, and how channels moderate the influence of self-disclosure on reciprocity and receiving support. By automatically measuring people's self-disclosure at scale, we found that members of cancer support groups revealed more negative self-disclosure in the public channels compared to the private channels. Although one's self-disclosure leads others to self-disclose and to provide support, these effects were generally stronger in the private channel. These channel effects probably occur because the public channels are the primary venue for support exchange, while the private channels are mainly used for follow-up conversations. We discuss theoretical and practical implications of our work.

In many teams, members play distinct roles, from leader to disrupter to social networker. Understanding the roles that contributors enact in micro-lending platforms is both psychologically and socially important for sustaining members' motivation and coordinating their joint efforts. Knowing what roles exist in these communities or which additional ones might be needed can also help teams function more effectively. In this paper, we identify social roles in Kiva.org, a peer-to-peer micro-funding platform, by utilizing members' lending behaviors, social network behaviors and communication behaviors to model their social roles at three levels. At the individual level, this method discovered active lenders who made many loans, early-bird lenders who made loans well before deadlines, and lurkers who rarely lent. At the topical level, our method differentiated those who had broad interests and lent to many causes from those who made loans only to borrowers in certain geographic regions or industry sectors. At the team level, the method revealed eight team-oriented functional roles such as encouragers, reminders, competitors, followers, and welcomers. To demonstrate the utility of the team roles, we used regression analysis to show how the distribution of social roles within teams influences the amount of money teams lent. Implications for identifying roles and understanding their contributions to teams are discussed.

Check-in data from social networks provide researchers a unique opportunity to model human dynamics at scale. However, it is unclear how indicative these check-in traces are of real human mobility. Prior work showed that significant amounts of Foursquare check-ins did not match with the physical mobility patterns of users, and suggested that misrepresented check-ins were incentivized by external rewards provided by the system. In this paper, our goal is to understand the root cause of inaccurate check-in data, by studying the validity of check-in traces in social media platforms without external rewards for check-ins. We conduct a data-driven analysis using an empirical check-in data trace of more than 276,000 users from WeChat Moments, with matching traces of their physical mobility. We develop a set of hypotheses on the underlying motivations behind people's inaccurate check-ins, and validate them using a detailed survey study. Our analysis reveals that there are surprisingly high amount of inaccurate check-ins even in the absence of rewards: 43% of total check-ins are inaccurate and 61% of survey participants report they have misrepresented their check-ins. We also find that inaccurate check-ins are often a result of user interface design as well as for convenience, self-advertisement and self-presentation.

For online communities to be successful, they must retain an adequate number of members who contribute to the community. The amount and type of communication members receive can play an important role in generating and sustaining members' commitment to it. However, the communication that members find valuable may change with their tenure in the community. This paper examines how the communication members receive in an health-support community influences their commitment and how this influence changes with their tenure in the community. Commitment was operationalized with three measures: self-reported attachment, continued participation in the community, and responding to others. Results show that receiving communication was generally associated with increased commitment across the three measures, with its impact increasing with members' tenure. However, the average amount of informational and emotional support members received per message was associated with decreased commitment. Results have implications for interventions to encourage members' commitment to their communities throughout their history in the community.

Most studies on human editing focus merely on syntactic revision operations, failing to capture the intentions behind revision changes, which are essential for facilitating the single and collaborative writing process.In this work, we develop in collaboration with Wikipedia editors a 13-category taxonomy of the semantic intention behind edits in Wikipedia articles.Using labeled article edits, we build a computational classifier of intentions that achieved a micro-averaged F1 score of 0.621.We use this model to investigate edit intention effectiveness: how different types of edits predict the retention of newcomers and changes in the quality of articles, two key concerns for Wikipedia today.Our analysis shows that the types of edits that users make in their first session predict their subsequent survival as Wikipedia editors, and articles in different stages need different types of edits.

Dual processing theories in psychology suggest that people process persuasive requests by assessing the quality of arguments (systematic processing) or by relying on heuristic rules (heuristic processing). However, the factors that act as systematic and heuristic processing cues and affect the success of persuasion have not been adequately described in social lending contexts. This research examines the effectiveness of systematic and heuristic cues in persuasive requests in Kiva lending teams intended to convince members to donate. An analysis of 88,596 requests exchanged in 1,610 teams shows that certain heuristic processing cues (e.g., liking between requesters and potential lenders, advocates low authority in their teams and the importance of the team to the lender) strongly predicted whether lenders would contribute to a loan request. In contrast, cues that required systematic processing are less influential. We also found that behavioral cues are more important than verbal ones. We discuss the theoretical and practical implications of our work.

Zichao Yang, Diyi Yang, Chris Dyer, Xiaodong He, Alex Smola, Eduard Hovy. Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2016.

Diyi Yang, Miaomiao Wen, Carolyn Rosé. Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2015.

We propose a novel data augmentation approach to enhance computational behavioral analysis using social media text.In particular, we collect a Twitter corpus of the descriptions of annoying behaviors using the #petpeeve hashtags.In the qualitative analysis, we study the language use in these tweets, with a special focus on the fine-grained categories and the geographic variation of the language.In quantitative analysis, we show that lexical and syntactic features are useful for automatic categorization of annoying behaviors, and frame-semantic features further boost the performance; that leveraging large lexical embeddings to create additional training instances significantly improves the lexical model; and incorporating frame-semantic embedding achieves the best overall performance.* We understand that many people find long titles annoying, so we intentionally use a very long one to help people understand what "pet peeve" means.

Humor is an essential component in personal communication. How to create computational models to discover the structures behind humor, recognize humor and even extract humor anchors remains a challenge. In this work, we first identify several semantic structures behind humor and design sets of features for each structure, and next employ a com-putational approach to recognize humor. Furthermore, we develop a simple and effective method to extract anchors that enable humor in a sentence. Experiments conducted on two datasets demonstrate that our humor recognizer is effective in automatically distinguishing between humorous and non-humorous texts and our extracted humor anchors correlate quite well with human annotations. 1

Thousands of students enroll in Massive Open Online Courses~(MOOCs) to seek opportunities for learning and self-improvement. However, the learning process often involves struggles with confusion, which may have an adverse effect on the course participation experience, leading to dropout along the way. In this paper, we quantify that effect. We describe a classification model using discussion forum behavior and clickstream data to automatically identify posts that express confusion. We then apply survival analysis to quantify the impact of confusion on student dropout. The results demonstrate that the more confusion students express or are exposed to, the lower the probability of their retention. Receiving support and resolution of confusion helps mitigate this effect. We explore the differential effects of confusion expressed in different contexts and related to different aspects of courses. We conclude with implications for design of interventions towards improving the retention of students in MOOCs.

This paper studies how to incorporate the external word correlation knowledge to improve the coherence of topic modeling.Existing topic models assume words are generated independently and lack the mechanism to utilize the rich similarity relationships among words to learn coherent topics.To solve this problem, we build a Markov Random Field (MRF) regularized Latent Dirichlet Allocation (LDA) model, which defines a MRF on the latent topic layer of LDA to encourage words labeled as similar to share the same topic label.Under our model, the topic assignment of each word is not independent, but rather affected by the topic labels of its correlated words.Similar words have better chance to be put into the same topic due to the regularization of MRF, hence the coherence of topics can be boosted.In addition, our model can accommodate the subtlety that whether two words are similar depends on which topic they appear in, which allows word with multiple senses to be put into different topics properly.We derive a variational inference method to infer the posterior probabilities and learn model parameters and present techniques to deal with the hardto-compute partition function in MRF.Experiments on two datasets demonstrate the effectiveness of our model.

A recent area in which recommender systems have shown their value is in online discussion forums and question-answer sites. Earlier work in this space has focused on the problem of matching participants to opportunities but has not adequately addressed the problem that in these social contexts, multiple dimensions of constraints must be satisfied, including limitations on capacity and minimal requirements for expertise. In this work, we propose such a constrained question recommendation problem with load balance constraints in discussion forums and use flow based model to generate the optimal solution. In particular, to address the introduced computation complexity, we investigate the concept of submodularity of the objective function and propose a specific submodular method to give an approximated solution. We present experiments conducted on two Massive Open Online Course (MOOC) discussion forum datasets, and demonstrate the effectiveness and efficiency of our submodular method in solving constrained question recommendation tasks.

In this paper, we explore student dropout behavior in a Massively Open Online Course (MOOC). We use a survival model to measure the impact of three social factors that make predictions about attrition along the way for students who have participated in the course discussion forum.

Massive Open Online Courses (MOOCs) have experienced a recent boom in interest. Problems students struggle with in the discussion forums, such as difficultly in finding interesting discussion opportunities or attracting helpers to address posted problems, provide new opportunities for recommender systems. In contrast to traditional product recommendation, question recommendation in discussion forums should simultaneously consider constraints on both students and questions. These considerations include (1) Load Balancing - students should not be over-burdened with too many requests; and (2) Expertise Matching - students should not be requested to address problems they are not capable of addressing. In this work, we formulate a novel constrained question recommendation problem to address the above considerations. We design a context-aware matrix factorization model to predict students' preferences over questions, then build a max cost flow model to manage the constraints. Experimental results conducted on three MOOC datasets demonstrate that our method significantly outperforms baseline methods in optimizing overall forum welfare, and in predicting which specific questions students might be interested in.

We study how an advertiser changes his/her bid prices in sponsored search, by modeling his/her rationality. Predicting the bid changes of advertisers with respect to their campaign performances is a key capability of search engines, since it can be used to improve the offline evaluation of new advertising technologies and the forecast of future revenue of the search engine. Previous work on advertiser behavior modeling heavily relies on the assumption of perfect advertiser rationality; however, in most cases, this assumption does not hold in practice. Advertisers may be unwilling, incapable, and/or constrained to achieve their best response. In this paper, we explicitly model these limitations in the rationality of advertisers, and build a probabilistic advertiser behavior model from the perspective of a search engine. We then use the expected payoff to define the objective function for an advertiser to optimize given his/her limited rationality. By solving the optimization problem with Monte Carlo, we get a prediction of mixed bid strategy for each advertiser in the next period of time. We examine the effectiveness of our model both directly using real historical bids and indirectly using revenue prediction and click number prediction. Our experimental results based on the sponsored search logs from a commercial search engine show that the proposed model can provide a more accurate prediction of advertiser bid behaviors than several baseline methods.

Digital music has experienced a quite fascinating transformation during the past decades. Thousands of people share or distribute their music collections on the Internet, resulting in an explosive increase of information and more user dependence on automatic recommender systems. Though there are many techniques such as collaborative filtering, most approaches focus mainly on users' global behaviors, neglecting local actions and the specific properties of music. In this paper, we propose a simple and effective local implicit feedback model mining users' local preferences to get better recommendation performance in both rating and ranking prediction. Moreover, we design an efficient training algorithm to speed up the updating procedure, and give a method to find the most appropriate time granularity to assist the performance. We conduct various experiments to evaluate the performance of this model, which show that it outperforms baseline model significantly. Integration with existing temporal models achieves a great improvement compared to the reported best single model for Yahoo! Music.

Recently, recommender systems have fascinated researchers and benefited a variety of people's online activities, enabling users to survive the explosive web information. Traditional collaborative filtering techniques handle the general recommendation well. However, most such approaches usually focus on long term preferences. To discover more short term factors influencing people's decisions, we propose a short term preferences model, implemented with implicit user feedback. We conduct experiments comparing the performances of different short term models, which show that our model outperforms significantly compared to those long term models.

Serendipitous recommendation has benefitted both e-retailers and users. It tends to suggest items which are both unexpected and useful to users. These items are not only profitable to the retailers but also surprisingly suitable to consumers' tastes. However, due to the imbalance in observed data for popular and tail items, existing collaborative filtering methods fail to give satisfactory serendipitous recommendations. To solve this problem, we propose a simple and effective method, called serendipitous personalized ranking. The experimental results demonstrate that our method significantly improves both accuracy and serendipity for top-N recommendation compared to traditional personalized ranking methods in various settings.