Cross-cultural communication often results in misaligned norms and expectations, leading to misunderstandings or harm.As the internet increasingly facilitates cross-cultural communication online, such misalignments also increase.However, there is an opportunity to use Large Language Models (LLMs) to detect such misunderstandings and assist in addressing them.To that end, this study investigates whether cross-cultural norm violations can be detected and mitigated using popular LLMs.Using a set of carefully constructed cross-cultural communication scenarios, half of which present norm violations, we test the ability of OpenAI's GPT-4 Omni (GPT-4o) model to identify cross-cultural communication norm violations.We find that GPT-4o classification accuracy varies by the stated age, gender, and nationality of the communicators described in the scenarios, suggesting a lack of fairness and a potential cultural gap in GPT-4o's detection.

Procedural content generation (PCG) in video games offers unprecedented opportunities for customization and user engagement. Working within the specialized context of role-playing games (RPGs), we introduce a novel framework for quest and dialogue generation that places the player at the core of the generative process. Drawing on a hand-crafted knowledge base, our method grounds generated content with in-game context while simultaneously employing a large-scale language model to create fluent, unique, accompanying dialogue. Through human evaluation, we confirm that quests generated using this method can approach the performance of hand-crafted quests in terms of fluency, coherence, novelty, and creativity; demonstrate the enhancement to the player experience provided by greater dynamism; and provide a novel, automated metric for the relevance between quest and dialogue. We view our contribution as a critical step toward dynamic, co-creative narrative frameworks in which humans and AI systems jointly collaborate to create unique and user-specific playable experiences.

We present an exploration of cultural norms surrounding online disclosure of information about one's interpersonal relationships (such as information about family members, colleagues, friends, or lovers) on Twitter. The literature identifies the cultural dimension of individualism versus collectivism as being a major determinant of offline communication differences in terms of emotion, topic, and content disclosed. We decided to study whether such differences also occur online in context of Twitter when comparing tweets posted in an individualistic (U.S.) versus a collectivist (India) society. We collected more than 2 million tweets posted in the U.S. and India over a 3 month period which contain interpersonal relationship keywords. A card-sort study was used to develop this culturally-sensitive saturated taxonomy of keywords that represent interpersonal relationships (e.g., ma, mom, mother). Then we developed a high-accuracy interpersonal disclosure detector based on dependency-parsing (F1-score: 86%) to identify when the words refer to a personal relationship of the poster (e.g., "my mom" as opposed to "a mom"). This allowed us to identify the 400K+ tweets in our data set which actually disclose information about the poster's interpersonal relationships. We used a mixed methods approach to analyze these tweets (e.g., comparing the amount of joy expressed about one's family) and found differences in emotion, topic, and content disclosed between tweets from the U.S. versus India. Our analysis also reveals how a combination of qualitative and quantitative methods are needed to uncover these differences; Using just one or the other can be misleading. This study extends the prior literature on Multi-Party Privacy and provides guidance for researchers and designers of culturally-sensitive systems.

Large transformer-based language models have achieved incredible success at various tasks which require narrative comprehension, including story completion, answering questions about stories, and generating stories ex nihilo. However, due to the limitations of finite context windows, these language models struggle to produce or understand stories longer than several thousand tokens. In order to mitigate the document length limitations that come with finite context windows, we introduce a novel architecture that augments story processing with an external dynamic knowledge graph. In contrast to static commonsense knowledge graphs which hold information about the real world, these dynamic knowledge graphs reflect facts extracted from the story being processed. Our architecture uses these knowledge graphs to create information-rich prompts which better facilitate story comprehension than prompts composed only of story text. We apply our architecture to the tasks of question answering and story completion. To complement this line of research, we introduce two long-form question answering tasks, LF-SQuAD and LF-QUOREF, in which the document length exceeds the size of the language model's context window, and introduce a story completion evaluation method that bypasses the stochastic nature of language model generation. We demonstrate broad improvement over typical prompt formulation methods for both question answering and story completion using GPT-2, GPT-3 and XLNet.

Taylor Sorensen, Joshua Robinson, Christopher Rytting, Alexander Shaw, Kyle Rogers, Alexia Delorey, Mahmoud Khalil, Nancy Fulda, David Wingate. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2022.

Many first-person shooters feature non-player characters (NPCs) that work alongside the player. Interfacing with these NPCs can add unnecessary complication to a game and steepen the learning curve for new players. Recent improvements in automated voice recognition and language representation have set the stage for more immersive methods of interfacing with NPCs through player speech. In this paper, we present several promising methods of classifying user utterances to extract predefined commands from unstructured speech. This framework facilitates a more flexible interface than has been used in past speech-controlled games. We also show how our methods effectively leverage small sets of example data to outperform existing industrial utterance classification systems.

It is notoriously difficult to control the behavior of artificial neural networks such as generative neural language models. We recast the problem of controlling natural language generation as that of learning to interface with a pretrained language model, just as Application Programming Interfaces (APIs) control the behavior of programs by altering hyperparameters. In this new paradigm, a specialized neural network (called a Neural Programming Interface or NPI) learns to interface with a pretrained language model by manipulating the hidden activations of the pretrained model to produce desired outputs. Importantly, no permanent changes are made to the weights of the original model, allowing us to re-purpose pretrained models for new tasks without overwriting any aspect of the language model. We also contribute a new data set construction algorithm and GAN-inspired loss function that allows us to train NPI models to control outputs of autoregressive transformers. In experiments against other state-of-the-art approaches, we demonstrate the efficacy of our methods using OpenAI’s GPT-2 model, successfully controlling noun selection, topic aversion, offensive speech filtering, and other aspects of language while largely maintaining the controlled model's fluency under deterministic settings.

Autonomous agents must often detect affordances: the set of behaviors enabled by a situation. Affordance extraction is particularly helpful in domains with large action spaces, allowing the agent to prune its search space by avoiding futile behaviors. This paper presents a method for affordance extraction via word embeddings trained on a tagged Wikipedia corpus. The resulting word vectors are treated as a common knowledge database which can be queried using linear algebra. We apply this method to a reinforcement learning agent in a text-only environment and show that affordance-based action selection improves performance in most cases. Our method increases the computational complexity of each learning step but significantly reduces the total number of steps needed. In addition, the agent's action selections begin to resemble those a human would choose.

Groups of reinforcement learning agents interacting in a common environment often fail to learn optimal behaviors. Poor performance is particularly common in environments where agents must coordinate with each other to receive rewards and where failed coordination attempts are penalized. This paper studies the effectiveness of the dynamic joint action perception (DJAP) algorithm on a grid-world rendezvous task with this characteristic. The effects of learning rate, exploration strategy, and training time on algorithm effectiveness are discussed. An analysis of the types of tasks for which DJAP learning is appropriate is also presented.

We present an equilibrium selection algorithm for reinforcement learning agents that incrementally adjusts the probability of executing each action based on the desirability of the outcome obtained in the last time step. The algorithm assumes that at least one coordination equilibrium exists and requires that the agents have a heuristic for determining whether or not the equilibrium was obtained. In deterministic environments with one or more strict coordination equilibria, the algorithm learns to play an optimal equilibrium as long as the heuristic is accurate. Empirical data demonstrate that the algorithm is also effective in stochastic environments and is able to learn good joint policies when the heuristic's parameters are estimated during learning, rather than known in advance.