Geological CO2 storage (GCS) involves injecting captured CO2 into deep subsurface formations to support climate goals. The effective management of GCS relies on adaptive injection planning to dynamically control injection rates and well pressures to balance both storage safety and efficiency. Prior literature, including numerical optimization methods and surrogate-optimization methods, is limited by real-world GCS requirements of smooth state transitions and goal-directed planning within limited time. To address these limitations, we propose a Brownian Bridge–augmented framework for surrogate simulation and injection planning in GCS and develop two insights (i) Brownian bridge as smooth state regularizer for better surrogate simulator; (ii) Brownian bridge as goal-time-conditioned planning guidance for better injection planning. Our method has three stages: (i) learning deep Brownian bridge representations with contrastive and reconstructive losses from historical reservoir and utility trajectories, (ii) incorporating Brownian bridge-based next state interpolation for simulator regularization (iii) guiding injection planning with Brownian utility-conditioned trajectories to generate high-quality injection plans. Experimental results across multiple datasets collected from diverse GCS settings demonstrate that our framework consistently improves simulation fidelity and planning effectiveness while maintaining low computational overhead.

Reasoning is a key component of language understanding in Large Language Models. While Chain-of-Thought prompting enhances performance via explicit intermediate steps, it suffers from sufficient token overhead and a fixed reasoning trajectory, preventing step-wise refinement. Recent advances in latent reasoning address these limitations by refining internal reasoning processes directly in the model’s latent space, without producing explicit outputs. However, a key challenge remains: how to effectively update reasoning embeddings during post-training to guide the model toward more accurate solutions. To overcome this challenge, we propose a lightweight post-training framework that refines latent reasoning trajectories using two novel strategies: (1) Contrastive reasoning feedback, which compares reasoning embeddings against strong and weak baselines to infer effective update directions via embedding enhancement; (2) Residual embedding refinement, which stabilizes updates by progressively integrating current and historical gradients, enabling fast yet controlled convergence. Extensive experiments and case studies are conducted on five reasoning benchmarks to demonstrate the effectiveness of the proposed framework. Notably, a +5% accuracy gain on MathQA without additional training.

Modern data-driven applications generate vast, complex data that often contain irrelevant, redundant, or noisy features. The task involves selecting the optimal subset of features from a dataset by removing redundant and irrelevant ones to enhance downstream performance. Using reinforcement learning (RL) for feature selection is ideal for data-centric tasks due to its interactive nature, which dynamically adapts to evolving data environments. However, high initial exploration variability, reliance on downstream tasks, and the need to control the false discovery rate bring challenges. To address these issues, we introduce a knockoff-guided RL framework, which uses pseudo-features to control early-stage randomness and matrix reconstruction for unsupervised reward. Our approach is a single-agent RL approach with unsupervised rewards that utilize knockoff information and matrix reconstruction to enhance feature selection without relying on labeled data. The framework integrates three components: the knockoff information to control exploration variability, decision network pre-training to guide RL policy, and matrix reconstruction to guide unsupervised rewards. Extensive experiments are conducted on various datasets across different task types, including classification and regression, demonstrating the superiority of our framework. Codes are available here.

Tabular data is one of the most widely used formats across industries, driving critical applications in areas such as finance, healthcare, and marketing. In the era of data-centric AI, improving data quality and representation has become essential for enhancing model performance, particularly in applications centered around tabular data. This survey examines the key aspects of tabular data-centric AI, emphasizing feature selection and feature generation as essential techniques for data space refinement. We provide a systematic review of feature selection methods, which identify and retain the most relevant data attributes, and feature generation approaches, which create new features to simplify the capture of complex data patterns. This survey offers a comprehensive overview of current methodologies through an analysis of recent advancements, practical applications, and the strengths and limitations of these techniques. Finally, we outline open challenges and suggest future perspectives to inspire continued innovation in this field.

Feature transformation involves generating a new set of features from the original dataset to enhance the data's utility. In certain domains like material performance screening, dimensionality is large and collecting labels is expensive and lengthy. It highly necessitates transforming feature spaces efficiently and without supervision to enhance data readiness and AI utility. However, existing methods fall short in efficient navigation of a vast space of feature combinations, and are mostly designed for supervised settings. To fill this gap, our unique perspective is to leverage a generator-critic duet-play teaming framework using LLM agents and in-context learning to derive pseudo-supervision from unsupervised data. The framework consists of three interconnected steps: (1) Critic agent diagnoses data to generate actionable advice, (2) Generator agent produces tokenized feature transformations guided by the critic's advice, and (3) Iterative refinement ensures continuous improvement through feedback between agents. The generator-critic framework can be generalized to human-agent collaborative generation, by replacing the critic agent with human experts. Extensive experiments demonstrate that the proposed framework outperforms even supervised baselines in feature transformation efficiency, robustness, and practical applicability across diverse datasets. Our code is publicly available at https://github.com/NanxuGong/LPFG.

The global COVID-19 pandemic underscored major deficiencies in traditional healthcare systems, hastening the advancement of online medical services, especially in medical triage and consultation. However, existing studies face two main challenges. First, the scarcity of large-scale, publicly available, domain-specific medical datasets due to privacy concerns, with current datasets being small and limited to a few diseases, limiting the effectiveness of triage methods based on Pre-trained Language Models (PLMs). Second, existing methods lack medical knowledge and struggle to accurately understand professional terms and expressions in patient-doctor consultations. To overcome these obstacles, we construct the Large-scale Chinese Medical Dialogue Corpora (LCMDC), thereby addressing the data shortage in this field. Moreover, we further propose a novel triage system that combines BERT-based supervised learning with prompt learning, as well as a GPT-based medical consultation model. To enhance domain knowledge acquisition, we pre-trained PLMs using our self-constructed background corpus. Experimental results on the LCMDC demonstrate the efficacy of our proposed systems. We further implement a voice-based medical dialogue system that enables real-time interaction between users and our models. Our corpora and codes are available at this link.

Xinyuan Wang, Yanchi Liu, Wei Cheng, Xujiang Zhao, Zhengzhang Chen, Wenchao Yu, Yanjie Fu, Haifeng Chen. 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.

Feature Transformation (FT) crafts new features from original ones via mathematical operations to enhance dataset expressiveness for downstream models. However, existing FT methods exhibit critical limitations: discrete search struggles with enormous combinatorial spaces, impeding practical use; and continuous search, being highly sensitive to initialization and step sizes, often becomes trapped in local optima, restricting global exploration. To overcome these limitations, DIFFT redefines FT as a reward-guided generative task. It first learns a compact and expressive latent space for feature sets using a Variational Auto-Encoder (VAE). A Latent Diffusion Model (LDM) then navigates this space to generate high-quality feature embeddings, its trajectory guided by a performance evaluator towards task-specific optima. This synthesis of global distribution learning (from LDM) and targeted optimization (reward guidance) produces potent embeddings, which a novel semi-autoregressive decoder efficiently converts into structured, discrete features, preserving intra-feature dependencies while allowing parallel inter-feature generation. Extensive experiments on 14 benchmark datasets show DIFFT consistently outperforms state-of-the-art baselines in predictive accuracy and robustness, with significantly lower training and inference times.

Graph recommendation methods, representing a connected interaction perspective, reformulate user–item interactions as graphs to leverage graph structure and topology to recommend and have proved practical effectiveness at scale. Large language models (LLMs), representing a textual generative perspective, excel at modeling user languages, understanding behavioral contexts, capturing user–item semantic relationships, analyzing textual sentiments, and generating coherent and contextually relevant texts as recommendations. However, there is a gap between the connected graph perspective and the text generation perspective as the task formulations are different. A research question arises: how can we effectively integrate the two perspectives for more personalized RecSys? To fill this gap, we propose to incorporate graph-edge information into LLMs via prompt and attention innovations. We reformulate recommendations as a probabilistic generative problem using prompts. We develop a framework to incorporate graph edge information from the prompt and attention mechanisms for graph-structured LLM recommendations. We develop a new prompt design that brings in both first-order and second-order graph relationships; we devise an improved LLM attention mechanism to embed direct the spatial and connectivity information of edges. Our evaluation of real-world datasets demonstrates the framework’s ability to understand connectivity information in graph data and to improve the relevance and quality of recommendation results. Our code is released at: https://github.com/anord-wang/LLM4REC.git .

Medical triage and diagnosis systems are crucial in alleviating the severe burden on healthcare systems, providing great assistance and convenience to patients with limited medical knowledge. Most current studies, however, struggle to effectively learn speaker-aware information in patient-doctor dialogues. Besides, the professional terms and disease-related expressions make it difficult for traditional methods to understand medical knowledge. To solve these challenges, we propose a novel Speaker-Aware Dual Encoding (SADE) model, which employs two encoders of identical structure to separately represent patient consultations and doctor diagnoses. To grasp medical knowledge, we introduce DiagBERT, pre-trained on numerous domain-specific texts to represent input utterances. We then integrate BiLSTM and Dendrite Network into each encoder to capture sequential and interactional semantics. Moreover, we have built two large-scale Triage and Diagnosis datasets to overcome the scarcity of public medical corpora. Extensive experimental results on these datasets indicate that SADE outperforms other baseline models.

Financial reports serve as crucial resources for investors and researchers, providing analysts’ assessments of stocks that play a vital role in stock market applications. However, detecting analysts’ opinions and sentiments in financial reports is challenging. First, the formal and professional language used in these reports makes it difficult for previous methods to comprehend domain-specific knowledge. Second, financial reports often adopt lengthy and elaborate expressions to convey rich semantics, which exposes the existing methods to contextual information loss, especially on long-term dependencies. To address these problems, we propose a summary-enhanced hierarchical framework (SEHF), which leverages summary information to enhance financial report sentiment analysis. Our framework incorporates financial bidirectional and auto-regressive transformer (FinBART), equipped with extended position encoding to summarize lengthy report articles and capture long-range interactions. To mitigate information loss, we initially divide each report into segments and then propose the hierarchical analyst sentiment representation network (ASRN), which utilizes financial bidirectional encoder representation from transformer (FinBERT), bidirectional long short-term memory (BiLSTM)-Attention, and dendrite (DD) network to fuse information in the generated summary and report segments. Notably, FinBART and FinBERT are pretrained on large-scale financial corpora to effectively understand professional expressions. Furthermore, we construct a new dataset large-scale Chinese financial report (LCFR) for the lack of supervised datasets. Experimental results on LCFR and a benchmark dataset show that SEHF significantly outperforms state-of-the-art (SOTA) baselines, and the ablation study highlights the effectiveness of aggregating sentiment information in the summary and report segments.

BACKGROUND: The zone adjacent to a transcription start site (TSS), namely, the promoter, is primarily involved in the process of DNA transcription initiation and regulation. As a result, proper promoter identification is critical for further understanding the mechanism of the networks controlling genomic regulation. A number of methodologies for the identification of promoters have been proposed. Nonetheless, due to the great heterogeneity existing in promoters, the results of these procedures are still unsatisfactory. In order to establish additional discriminative characteristics and properly recognize promoters, we developed the hybrid model for promoter identification (HMPI), a hybrid deep learning model that can characterize both the native sequences of promoters and the morphological outline of promoters at the same time. We developed the HMPI to combine a method called the PSFN (promoter sequence features network), which characterizes native promoter sequences and deduces sequence features, with a technique referred to as the DSPN (deep structural profiles network), which is specially structured to model the promoters in terms of their structural profile and to deduce their structural attributes. RESULTS: The HMPI was applied to human, plant and Escherichia coli K-12 strain datasets, and the findings showed that the HMPI was successful at extracting the features of the promoter while greatly enhancing the promoter identification performance. In addition, after the improvements of synthetic sampling, transfer learning and label smoothing regularization, the improved HMPI models achieved good results in identifying subtypes of promoters on prokaryotic promoter datasets. CONCLUSIONS: The results showed that the HMPI was successful at extracting the features of promoters while greatly enhancing the performance of identifying promoters on both eukaryotic and prokaryotic datasets, and the improved HMPI models are good at identifying subtypes of promoters on prokaryotic promoter datasets. The HMPI is additionally adaptable to different biological functional sequences, allowing for the addition of new features or models.