Peer Revenue Fully funded

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

Computer Science Computation and Language

cs.CL Watch Computation and Language

Wednesday 24 December

Computation and Language

Opportunities and Challenges of Natural Language Processing for Low-Resource Senegalese Languages in Social Science Research

Derguene Mbaye, Tatiana D. P. Mbengue, Madoune R. Seye et al. (+7 more)

Natural Language Processing (NLP) is rapidly transforming research methodologies across disciplines, yet African languages remain largely underrepresented in this technological shift. This paper provides the first comprehensive overview of NLP progress and challenges for the six national languages officially recognized by the Senegalese Constitution: Wolof, Pulaar, Sérère, Diola, Mandingue, and Soninké. We synthesize linguistic, socio-technical, and infrastructural factors that shape their digital readiness and identify gaps in data, tools, and benchmarks. Building on existing initiatives and research works, we analyze ongoing efforts in various tasks, covering both text and speech modalities. We also provide a centralized GitHub repository that compiles publicly accessible resources for a range of NLP tasks across these languages, designed to facilitate collaboration and reproducibility. A special focus is devoted to the application of NLP to the social sciences, where multilingual transcription, translation, and retrieval pipelines can significantly enhance the efficiency and inclusiveness of field research. The paper concludes by outlining a roadmap toward sustainable, community-centered NLP ecosystems for Senegalese languages, emphasizing ethical data governance, open resources, and interdisciplinary collaboration.
Read the full abstract

All authors Derguene Mbaye, Tatiana D. P. Mbengue, Madoune R. Seye, Moussa Diallo, Mamadou L. Ndiaye, Dimitri S. Adjanohoun, Cheikh S. Wade, Djiby Sow, Jean-Claude B. Munyaka, Jerome Chenal

Computation and Language

Introducing Axlerod: An LLM-based Chatbot for Assisting Independent Insurance Agents

Adam Bradley, John Hastings and Khandaker Mamun Ahmed

The insurance industry is undergoing a paradigm shift through the adoption of artificial intelligence (AI) technologies, particularly in the realm of intelligent conversational agents. Chatbots have evolved into sophisticated AI-driven systems capable of automating complex workflows, including policy recommendation and claims triage, while simultaneously enabling dynamic, context-aware user engagement. This paper presents the design, implementation, and empirical evaluation of Axlerod, an AI-powered conversational interface designed to improve the operational efficiency of independent insurance agents. Leveraging natural language processing (NLP), retrieval-augmented generation (RAG), and domain-specific knowledge integration, Axlerod demonstrates robust capabilities in parsing user intent, accessing structured policy databases, and delivering real-time, contextually relevant responses. Experimental results underscore Axlerod's effectiveness, achieving an overall accuracy of 93.18% in policy retrieval tasks while reducing the average search time by 2.42 seconds. This work contributes to the growing body of research on enterprise-grade AI applications in insurtech, with a particular focus on agent-assistive rather than consumer-facing architectures.
Read the full abstract

Computation and Language

Evaluating Novelty in AI-Generated Research Plans Using Multi-Workflow LLM Pipelines

Devesh Saraogi, Rohit Singhee and Dhruv Kumar

The integration of Large Language Models (LLMs) into the scientific ecosystem raises fundamental questions about the creativity and originality of AI-generated research. Recent work has identified “smart plagiarism” as a concern in single-step prompting approaches, where models reproduce existing ideas with terminological shifts. This paper investigates whether agentic workflows – multi-step systems employing iterative reasoning, evolutionary search, and recursive decomposition – can generate more novel and feasible research plans. We benchmark five reasoning architectures: Reflection-based iterative refinement, Sakana AI v2 evolutionary algorithms, Google Co-Scientist multi-agent framework, GPT Deep Research (GPT-5.1) recursive decomposition, and Gemini~3 Pro multimodal long-context pipeline. Using evaluations from thirty proposals each on novelty, feasibility, and impact, we find that decomposition-based and long-context workflows achieve mean novelty of 4.17/5, while reflection-based approaches score significantly lower (2.33/5). Results reveal varied performance across research domains, with high-performing workflows maintaining feasibility without sacrificing creativity. These findings support the view that carefully designed multi-stage agentic workflows can advance AI-assisted research ideation.
Read the full abstract

Computation and Language

LLM-Driven Preference Data Synthesis for Proactive Prediction of the Next User Utterance in Human-Machine Dialogue

Jinqiang Wang, Huansheng Ning, Jianguo Ding et al. (+3 more)

Proactively predicting a users next utterance in human-machine dialogue can streamline interaction and improve user experience. Existing commercial API-based solutions are subject to privacy concerns while deploying general-purpose LLMs locally remains computationally expensive. As such, training a compact, task-specific LLM provides a practical alternative. Although user simulator methods can predict a user's next utterance, they mainly imitate their speaking style rather than advancing the dialogue. Preference data synthesis has been investigated to generate data for proactive next utterance prediction and help align LLMs with user preferences. Yet existing methods lack the ability to explicitly model the intent reasoning that leads to the user's next utterance and to define and synthesize preference and non-preference reasoning processes for predicting the user's next utterance.To address these challenges, we propose ProUtt, an LLM-driven preference data synthesis method for proactive next utterance prediction. ProUtt converts dialogue history into an intent tree and explicitly models intent reasoning trajectories by predicting the next plausible path from both exploitation and exploration perspectives. It then constructs preference and non-preference reasoning processes by perturbing or revising intent tree paths at different future turns. Extensive evaluations using LLM-as-a-judge and human judgments demonstrate that ProUtt consistently outperforms existing data synthesis methods, user simulators, and commercial LLM APIs across four benchmark datasets. We release both the code and the synthesized datasets to facilitate future research.
Read the full abstract

All authors Jinqiang Wang, Huansheng Ning, Jianguo Ding, Tao Zhu, Liming Chen, Chris Nugent

Computation and Language

LLM_annotate: A Python package for annotating and analyzing fiction characters

Hannes Rosenbusch

LLM_annotate is a Python package for analyzing the personality of fiction characters with large language models. It standardizes workflows for annotating character behaviors in full texts (e.g., books and movie scripts), inferring character traits, and validating annotation/inference quality via a human-in-the-loop GUI. The package includes functions for text chunking, LLM-based annotation, character name disambiguation, quality scoring, and computation of character-level statistics and embeddings. Researchers can use any LLM, commercial, open-source, or custom, within LLM_annotate. Through tutorial examples using The Simpsons Movie and the novel Pride and Prejudice, I demonstrate the usage of the package for efficient and reproducible character analyses.
Read the full abstract

Computation and Language

Morality is Contextual: Learning Interpretable Moral Contexts from Human Data with Probabilistic Clustering and Large Language Models

Geoffroy Morlat, Marceau Nahon, Augustin Chartouny et al. (+3 more)

A key question in current AI alignment research is how to make AI algorithms learn moral values. Because human morality is highly context-dependent, actions are judged not only by their outcomes but by the context in which they occur. We present COMETH (Contextual Organization of Moral Evaluation from Textual Human inputs), a framework that integrates a probabilistic context learner with LLM-based semantic abstraction and human moral evaluations to model how context shapes the acceptability of ambiguous actions. We curate an empirically grounded dataset of 300 scenarios across six core actions relative to three moral rules (violating "Do not kill", "Do not deceive", and "Do not break the law") and collect ternary judgments (Blame/Neutral/Support) from N=101 participants. A preprocessing pipeline standardizes actions via an LLM filter and MiniLM embeddings with K-means, producing robust, reproducible core-action clusters. COMETH then learns action-specific moral contexts by clustering scenarios online from human judgment distributions using principled divergence criteria. To generalize and explain predictions, a Generalization module extracts concise, non-evaluative binary contextual features and learns feature weights in a transparent likelihood-based model. Empirically, COMETH roughly doubles alignment score with most human judgments relative to end-to-end LLM prompting (60% vs. 30% on average), while revealing which contextual features drive its predictions. The contributions are: (i) an empirically grounded moral-context dataset, (ii) a reproducible pipeline combining human judgments with model-based context learning and LLM semantics, and (iii) a more interpretable alternative than end-to-end LLMs for context-sensitive moral prediction and explanation.
Read the full abstract

All authors Geoffroy Morlat, Marceau Nahon, Augustin Chartouny, Raja Chatila, Ismael T. Freire, Mehdi Khamassi

Computation and Language

Teaching People LLM's Errors and Getting it Right

Nathan Stringham, Fateme Hashemi Chaleshtori, Xinyuan Yan et al. (+3 more)

People use large language models (LLMs) when they should not. This is partly because they see LLMs compose poems and answer intricate questions, so they understandably, but incorrectly, assume LLMs won't stumble on basic tasks like simple arithmetic. Prior work has tried to address this by clustering instance embeddings into regions where an LLM is likely to fail and automatically describing patterns in these regions. The found failure patterns are taught to users to mitigate their overreliance. Yet, this approach has not fully succeeded. In this analysis paper, we aim to understand why. We first examine whether the negative result stems from the absence of failure patterns. We group instances in two datasets by their meta-labels and evaluate an LLM's predictions on these groups. We then define criteria to flag groups that are sizable and where the LLM is error-prone, and find meta-label groups that meet these criteria. Their meta-labels are the LLM's failure patterns that could be taught to users, so they do exist. We next test whether prompting and embedding-based approaches can surface these known failures. Without this, users cannot be taught about them to reduce their overreliance. We find mixed results across methods, which could explain the negative result. Finally, we revisit the final metric that measures teaching effectiveness. We propose to assess a user's ability to effectively use the given failure patterns to anticipate when an LLM is error-prone. A user study shows a positive effect from teaching with this metric, unlike the human-AI team accuracy. Our findings show that teaching failure patterns could be a viable approach to mitigating overreliance, but success depends on better automated failure-discovery methods and using metrics like ours.
Read the full abstract

All authors Nathan Stringham, Fateme Hashemi Chaleshtori, Xinyuan Yan, Zhichao Xu, Bei Wang, Ana Marasović

Computation and Language

Optimizing Decoding Paths in Masked Diffusion Models by Quantifying Uncertainty

Ziyu Chen, Xinbei Jiang, Peng Sun et al. (+1 more)

Masked Diffusion Models (MDMs) offer flexible, non-autoregressive generation, but this freedom introduces a challenge: final output quality is highly sensitive to the decoding order. We are the first to formalize this issue, attributing the variability in output quality to the cumulative predictive uncertainty along a generative path. To quantify this uncertainty, we introduce Denoising Entropy, a computable metric that serves as an internal signal for evaluating generative process. Leveraging this metric, we propose two algorithms designed to optimize the decoding path: a post-hoc selection method and a real-time guidance strategy. Experiments demonstrate that our entropy-guided methods significantly improve generation quality, consistently boosting accuracy on challenging reasoning, planning, and code benchmarks. Our work establishes Denoising Entropy as a principled tool for understanding and controlling generation, effectively turning the uncertainty in MDMs from a liability into a key advantage for discovering high-quality solutions.
Read the full abstract

All authors Ziyu Chen, Xinbei Jiang, Peng Sun, Tao Lin

Computation and Language

C2LLM Technical Report: A New Frontier in Code Retrieval via Adaptive Cross-Attention Pooling

Jin Qin, Zihan Liao, Ziyin Zhang et al. (+3 more)

We present C2LLM - Contrastive Code Large Language Models, a family of code embedding models in both 0.5B and 7B sizes. Building upon Qwen-2.5-Coder backbones, C2LLM adopts a Pooling by Multihead Attention (PMA) module for generating sequence embedding from token embeddings, effectively 1) utilizing the LLM's causal representations acquired during pretraining, while also 2) being able to aggregate information from all tokens in the sequence, breaking the information bottleneck in EOS-based sequence embeddings, and 3) supporting flexible adaptation of embedding dimension, serving as an alternative to MRL. Trained on three million publicly available data, C2LLM models set new records on MTEB-Code among models of similar sizes, with C2LLM-7B ranking 1st on the overall leaderboard.
Read the full abstract

All authors Jin Qin, Zihan Liao, Ziyin Zhang, Hang Yu, Peng Di, Rui Wang

Computation and Language

Your Reasoning Benchmark May Not Test Reasoning: Revealing Perception Bottleneck in Abstract Reasoning Benchmarks

Xinhe Wang, Jin Huang, Xingjian Zhang et al. (+2 more)

Reasoning benchmarks such as the Abstraction and Reasoning Corpus (ARC) and ARC-AGI are widely used to assess progress in artificial intelligence and are often interpreted as probes of core, so-called “fluid” reasoning abilities. Despite their apparent simplicity for humans, these tasks remain challenging for frontier vision-language models (VLMs), a gap commonly attributed to deficiencies in machine reasoning. We challenge this interpretation and hypothesize that the gap arises primarily from limitations in visual perception rather than from shortcomings in inductive reasoning. To verify this hypothesis, we introduce a two-stage experimental pipeline that explicitly separates perception and reasoning. In the perception stage, each image is independently converted into a natural-language description, while in the reasoning stage a model induces and applies rules using these descriptions. This design prevents leakage of cross-image inductive signals and isolates reasoning from perception bottlenecks. Across three ARC-style datasets, Mini-ARC, ACRE, and Bongard-LOGO, we show that the perception capability is the dominant factor underlying the observed performance gap by comparing the two-stage pipeline with against standard end-to-end one-stage evaluation. Manual inspection of reasoning traces in the VLM outputs further reveals that approximately 80 percent of model failures stem from perception errors. Together, these results demonstrate that ARC-style benchmarks conflate perceptual and reasoning challenges and that observed performance gaps may overstate deficiencies in machine reasoning. Our findings underscore the need for evaluation protocols that disentangle perception from reasoning when assessing progress in machine intelligence.
Read the full abstract

All authors Xinhe Wang, Jin Huang, Xingjian Zhang, Tianhao Wang, Jiaqi W. Ma

Computation and Language · Cross-listed from cs.LG

Measuring all the noises of LLM Evals

Sida I. Wang

Applying well-established statistical methods effectively to LLM evals requires consideration of model correlations and prediction noises. Leading evals often ignore data noise, have low statistical power, or make unnecessary trade-offs. We define and measure three types of noise: 1) prediction noise from generating different answers on a given question, 2) data noise from sampling questions, and 3) their combined total noise following the law of total variance. To emphasize relative comparisons and gain statistical power, we propose all-pairs comparison, which applies paired comparison to all pairs of LLMs and measures all the noise components based on millions of question-level predictions across many evals and settings, revealing clear trends. First, most evals exhibit a characteristic and highly predictable total noise level across all model pairs. Second, paired prediction noise typically exceeds paired data noise, which means reducing prediction noise by averaging can significantly increase statistical power. By measuring all the noises together, we can assess eval results in context and make better empirical decisions.
Read the full abstract

Computation and Language

Parallel Token Prediction for Language Models

Felix Draxler, Justus Will, Farrin Marouf Sofian et al. (+3 more)

Autoregressive decoding in language models is inherently slow, generating only one token per forward pass. We propose Parallel Token Prediction (PTP), a general-purpose framework for predicting multiple tokens in a single model call. PTP moves the source of randomness from post-hoc sampling to random input variables, making future tokens deterministic functions of those inputs and thus jointly predictable in a single forward pass. We prove that a single PTP call can represent arbitrary dependencies between tokens. PTP is trained by distilling an existing model or through inverse autoregressive training without a teacher. Experimentally, PTP achieves a 2.4x speedup on a diverse-task speculative decoding benchmark. We provide code and checkpoints at https://github.com/mandt-lab/ptp.
Read the full abstract

All authors Felix Draxler, Justus Will, Farrin Marouf Sofian, Theofanis Karaletsos, Sameer Singh, Stephan Mandt

Computation and Language

SMART SLM: Structured Memory and Reasoning Transformer, A Small Language Model for Accurate Document Assistance

Divij Dudeja and Mayukha Pal

The user of Engineering Manuals (EM) finds it difficult to read EM s because they are long, have a dense format which includes written documents, step by step procedures, and standard parameter lists for engineering equipment. Off the shelf transformers, especially compact ones, treat this material as a flat stream of tokens. This approach leads to confident but incorrect numeric answers and forces the models to memorize separate facts inefficiently. SMART (Structured Memory and Reasoning Transformer) offers a different and practical solution to the above problem. SMART structures its processing by using a hierarchical approach, and is based upon three main job categories (1) A syntax-aware Fact Extractor (Grammarian) Tree LSTM which extracts facts as subject relation object relations from EM sentences (2) A compact indexed memory MANN (Memory Augmented Neural Network) that indexes these Rational Subject Relation Objects as 384 dimensional vectors that are associated with the source of the information, and (3) A 6 layer Transformer that learns to fuse the previously retrieved facts into its generated response. The entire SMART model utilizes 45.51M parameters, which is 64% less than GPT-2 (124M) and 69% less than BERT (133M), and it achieves a 21.3% higher accuracy than GPT-2, indicating that SMART fits the data better with the least amount of processing requirements. SMART employs dual modes of inference an indexed fast path for known documents (sub-second answer times) and an indexed dynamic path assisted by RAGs for new uploads (FAISS Top 20 results with memory severed at 64 slots). In real world deployment, this framework leads to more well supported results with reduced hallucinations than comparable small transformer models.
Read the full abstract

Computation and Language · Cross-listed from cs.IR

ReaSeq: Unleashing World Knowledge via Reasoning for Sequential Modeling

Jiakai Tang, Chuan Wang, Gaoming Yang et al. (+32 more)

Industrial recommender systems face two fundamental limitations under the log-driven paradigm: (1) knowledge poverty in ID-based item representations that causes brittle interest modeling under data sparsity, and (2) systemic blindness to beyond-log user interests that constrains model performance within platform boundaries. These limitations stem from an over-reliance on shallow interaction statistics and close-looped feedback while neglecting the rich world knowledge about product semantics and cross-domain behavioral patterns that Large Language Models have learned from vast corpora. To address these challenges, we introduce ReaSeq, a reasoning-enhanced framework that leverages world knowledge in Large Language Models to address both limitations through explicit and implicit reasoning. Specifically, ReaSeq employs explicit Chain-of-Thought reasoning via multi-agent collaboration to distill structured product knowledge into semantically enriched item representations, and latent reasoning via Diffusion Large Language Models to infer plausible beyond-log behaviors. Deployed on Taobao's ranking system serving hundreds of millions of users, ReaSeq achieves substantial gains: >6.0% in IPV and CTR, >2.9% in Orders, and >2.5% in GMV, validating the effectiveness of world-knowledge-enhanced reasoning over purely log-driven approaches.
Read the full abstract

All authors Jiakai Tang, Chuan Wang, Gaoming Yang, Han Wu, Jiahao Yu, Jian Wu, Jianwu Hu, Junjun Zheng, Longbin Li, Shuwen Xiao, Xiangheng Kong, Yeqiu Yang, Yuning Jiang, Ahjol Nurlanbek, Binbin Cao, Bo Zheng, Fangmei Zhu, Gaoming Zhou, Huimin Yi, Huiping Chu, Jin Huang, Jinzhe Shan, Kenan Cui, Longbin Li, Silu Zhou, Wen Chen, Xia Ming, Xiang Gao, Xin Yao, Xingyu Wen, Yan Zhang, Yiwen Hu, Yulin Wang, Ziheng Bao, Zongyuan Wu

Computation and Language

SpidR-Adapt: A Universal Speech Representation Model for Few-Shot Adaptation

Mahi Luthra, Jiayi Shen, Maxime Poli et al. (+14 more)

Human infants, with only a few hundred hours of speech exposure, acquire basic units of new languages, highlighting a striking efficiency gap compared to the data-hungry self-supervised speech models. To address this gap, this paper introduces SpidR-Adapt for rapid adaptation of speech units to new languages using minimal unlabeled data. We cast such low-resource speech representation learning as a meta-learning problem and construct a multi-task adaptive pre-training (MAdaPT) protocol which formulates the adaptation process as a bi-level optimization framework. To enable scalable meta-training under this framework, we propose a novel heuristic solution, first-order bi-level optimization (FOBLO), avoiding heavy computation costs. Finally, we stabilize meta-training by using a robust initialization through interleaved supervision which alternates self-supervised and supervised objectives. Empirically, SpidR-Adapt achieves rapid gains in phonemic discriminability (ABX) and downstream spoken language modeling scores (sWUGGY, sBLIMP, tSC), surpassing in-domain toplines after training on less than 1h of target-language audio and delivering $100\times$ greater data efficiency than standard multi-task training. These findings highlight a practical, architecture-agnostic path toward biologically inspired, data-efficient representations. We open-source the training code and model checkpoints at https://github.com/facebookresearch/spidr-adapt.
Read the full abstract

All authors Mahi Luthra, Jiayi Shen, Maxime Poli, Angelo Ortiz, Yosuke Higuchi, Youssef Benchekroun, Martin Gleize, Charles-Eric Saint-James, Dongyan Lin, Phillip Rust, Angel Villar, Surya Parimi, Vanessa Stark, Rashel Moritz, Juan Pino, Yann LeCun, Emmanuel Dupoux

Computation and Language

ClarifyMT-Bench: Benchmarking and Improving Multi-Turn Clarification for Conversational Large Language Models

Sichun Luo, Yi Huang, Mukai Li et al. (+5 more)

Large language models (LLMs) are increasingly deployed as conversational assistants in open-domain, multi-turn settings, where users often provide incomplete or ambiguous information. However, existing LLM-focused clarification benchmarks primarily assume single-turn interactions or cooperative users, limiting their ability to evaluate clarification behavior in realistic settings. We introduce ClarifyMT-Bench, a benchmark for multi-turn clarification grounded in a five-dimensional ambiguity taxonomy and a set of six behaviorally diverse simulated user personas. Through a hybrid LLM-human pipeline, we construct 6,120 multi-turn dialogues capturing diverse ambiguity sources and interaction patterns. Evaluating ten representative LLMs uncovers a consistent under-clarification bias: LLMs tend to answer prematurely, and performance degrades as dialogue depth increases. To mitigate this, we propose ClarifyAgent, an agentic approach that decomposes clarification into perception, forecasting, tracking, and planning, substantially improving robustness across ambiguity conditions. ClarifyMT-Bench establishes a reproducible foundation for studying when LLMs should ask, when they should answer, and how to navigate ambiguity in real-world human-LLM interactions.
Read the full abstract

All authors Sichun Luo, Yi Huang, Mukai Li, Shichang Meng, Fengyuan Liu, Zefa Hu, Junlan Feng, Qi Liu

Computation and Language · Cross-listed from cs.AI

Beyond Context: Large Language Models' Failure to Grasp Users' Intent

Ahmed M. Hussain and Salahuddin Salahuddin

Current Large Language Models (LLMs) safety approaches focus on explicitly harmful content while overlooking a critical vulnerability: the inability to understand context and recognize user intent. This creates exploitable vulnerabilities that malicious users can systematically leverage to circumvent safety mechanisms. We empirically evaluate multiple state-of-the-art LLMs, including ChatGPT, Claude, Gemini, and DeepSeek. Our analysis demonstrates the circumvention of reliable safety mechanisms through emotional framing, progressive revelation, and academic justification techniques. Notably, reasoning-enabled configurations amplified rather than mitigated the effectiveness of exploitation, increasing factual precision while failing to interrogate the underlying intent. The exception was Claude Opus 4.1, which prioritized intent detection over information provision in some use cases. This pattern reveals that current architectural designs create systematic vulnerabilities. These limitations require paradigmatic shifts toward contextual understanding and intent recognition as core safety capabilities rather than post-hoc protective mechanisms.
Read the full abstract

Computation and Language

Semi-Supervised Learning for Large Language Models Safety and Content Moderation

Eduard Stefan Dinuta, Iustin Sirbu and Traian Rebedea

Safety for Large Language Models (LLMs) has been an ongoing research focus since their emergence and is even more relevant nowadays with the increasing capacity of those models. Currently, there are several guardrails in place for all public LLMs and multiple proposed datasets for training safety classifiers. However, training these safety classifiers relies on large quantities of labeled data, which can be problematic to acquire, prone to labeling errors, or often include synthetic data. To address these issues, we suggest a different approach: utilizing semi-supervised learning techniques, which leverage both labeled and unlabeled data, to improve the performance on the safety task. We analyze the improvements that these techniques can offer for both prompts given to Large Language Models and the responses to those requests. Moreover, since augmentation is the central part of semi-supervised algorithms, we demonstrate the importance of using task-specific augmentations, which significantly increase the performance when compared to general-purpose augmentation techniques.
Read the full abstract

Computation and Language

Semantic Refinement with LLMs for Graph Representations

Safal Thapaliya, Zehong Wang, Jiazheng Li et al. (+3 more)

Graph-structured data exhibit substantial heterogeneity in where their predictive signals originate: in some domains, node-level semantics dominate, while in others, structural patterns play a central role. This structure-semantics heterogeneity implies that no graph learning model with a fixed inductive bias can generalize optimally across diverse graph domains. However, most existing methods address this challenge from the model side by incrementally injecting new inductive biases, which remains fundamentally limited given the open-ended diversity of real-world graphs. In this work, we take a data-centric perspective and treat node semantics as a task-adaptive variable. We propose a Graph-Exemplar-guided Semantic Refinement (GES) framework for graph representation learning which – unlike existing LLM-enhanced methods that generate node descriptions without graph context – leverages structurally and semantically similar nodes from the graph itself to guide semantic refinement. Specifically, a GNN is first trained to produce predictive states, which along with structural and semantic similarity are used to retrieve in-graph exemplars that inform an LLM in refining node descriptions. We evaluate our approach on both text-rich and text-free graphs. Results show consistent improvements on semantics-rich and structure-dominated graphs, demonstrating the effectiveness of data-centric semantic refinement under structure-semantics heterogeneity.
Read the full abstract

All authors Safal Thapaliya, Zehong Wang, Jiazheng Li, Ziming Li, Yanfang Ye, Chuxu Zhang

Computation and Language

Rethinking Supervised Fine-Tuning: Emphasizing Key Answer Tokens for Improved LLM Accuracy

Xiaofeng Shi, Qian Kou, Yuduo Li et al. (+1 more)

With the rapid advancement of Large Language Models (LLMs), the Chain-of-Thought (CoT) component has become significant for complex reasoning tasks. However, in conventional Supervised Fine-Tuning (SFT), the model could allocate disproportionately more attention to CoT sequences with excessive length. This reduces focus on the much shorter but essential Key portion-the final answer, whose correctness directly determines task success and evaluation quality. To address this limitation, we propose SFTKey, a two-stage training scheme. In the first stage, conventional SFT is applied to ensure proper output format, while in the second stage, only the Key portion is fine-tuned to improve accuracy. Extensive experiments across multiple benchmarks and model families demonstrate that SFTKey achieves an average accuracy improvement exceeding 5% over conventional SFT, while preserving the ability to generate correct formats. Overall, this study advances LLM fine-tuning by explicitly balancing CoT learning with additional optimization on answer-relevant tokens.
Read the full abstract

All authors Xiaofeng Shi, Qian Kou, Yuduo Li, Hua Zhou

Computation and Language

Distilling the Essence: Efficient Reasoning Distillation via Sequence Truncation

Wei-Rui Chen, Vignesh Kothapalli, Ata Fatahibaarzi et al. (+5 more)

Distilling the capabilities from a large reasoning model (LRM) to a smaller student model often involves training on substantial amounts of reasoning data. However, knowledge distillation (KD) over lengthy sequences with prompt (P), chain-of-thought (CoT), and answer (A) sections makes the process computationally expensive. In this work, we investigate how the allocation of supervision across different sections (P, CoT, A) affects student performance. Our analysis shows that selective KD over only the CoT tokens can be effective when the prompt and answer information is encompassed by it. Building on this insight, we establish a truncation protocol to quantify computation-quality tradeoffs as a function of sequence length. We observe that beyond a specific length, longer training sequences provide marginal returns for downstream performance but require substantially higher memory and FLOPs. To this end, training on only the first $50\%$ of tokens of every training sequence can retain, on average, $\approx91\%$ of full-sequence performance on math benchmarks while reducing training time, memory usage, and FLOPs by about $50\%$ each. Codes are available at https://github.com/weiruichen01/distilling-the-essence.
Read the full abstract

All authors Wei-Rui Chen, Vignesh Kothapalli, Ata Fatahibaarzi, Hejian Sang, Shao Tang, Qingquan Song, Zhipeng Wang, Muhammad Abdul-Mageed

Computation and Language

Automatic Replication of LLM Mistakes in Medical Conversations

Oleksii Proniakin, Diego Fajardo, Ruslan Nazarenko et al. (+1 more)

Large language models (LLMs) are increasingly evaluated in clinical settings using multi-dimensional rubrics which quantify reasoning quality, safety, and patient-centeredness. Yet, replicating specific mistakes in other LLM models is not straightforward and often requires manual effort. We introduce MedMistake, an automatic pipeline that extracts mistakes LLMs make in patient-doctor conversations and converts them into a benchmark of single-shot QA pairs. Our pipeline (1) creates complex, conversational data between an LLM patient and LLM doctor, (2) runs an evaluation with a committee of 2 LLM judges across a variety of dimensions and (3) creates simplified single-shot QA scenarios from those mistakes. We release MedMistake-All, a dataset of 3,390 single-shot QA pairs where GPT-5 and Gemini 2.5 Pro are currently failing to answer correctly, as judged by two LLM judges. We used medical experts to validate a subset of 211/3390 questions (MedMistake-Bench), which we used to run a final evaluation of 12 frontier LLMs: Claude Opus 4.5, Claude Sonnet 4.5, DeepSeek-Chat, Gemini 2.5 Pro, Gemini 3 Pro, GPT-4o, GPT-5, GPT-5.1, GPT-5.2, Grok 4, Grok 4.1, Mistral Large. We found that GPT models, Claude and Grok obtained the best performance on MedMistake-Bench. We release both the doctor-validated benchmark (MedMistake-Bench), as well as the full dataset (MedMistake-All) at https://huggingface.co/datasets/TheLumos/MedicalMistakeBenchmark.
Read the full abstract

All authors Oleksii Proniakin, Diego Fajardo, Ruslan Nazarenko, Razvan Marinescu

Computation and Language

Reflection Pretraining Enables Token-Level Self-Correction in Biological Sequence Models

Xiang Zhang, Jiaqi Wei, Yuejin Yang et al. (+8 more)

Chain-of-Thought (CoT) prompting has significantly advanced task-solving capabilities in natural language processing with large language models. Unlike standard prompting, CoT encourages the model to generate intermediate reasoning steps, non-answer tokens, that help guide the model toward more accurate final outputs. These intermediate steps enable more complex reasoning processes such as error correction, memory management, future planning, and self-reflection. However, applying CoT to non-natural language domains, such as protein and RNA language models, is not yet possible, primarily due to the limited expressiveness of their token spaces (e.g., amino acid tokens). In this work, we propose and define the concept of language expressiveness: the ability of a given language, using its tokens and grammar, to encode information. We show that the limited expressiveness of protein language severely restricts the applicability of CoT-style reasoning. To overcome this, we introduce reflection pretraining, for the first time in a biological sequence model, which enables the model to engage in intermediate reasoning through the generation of auxiliary "thinking tokens" beyond simple answer tokens. Theoretically, we demonstrate that our augmented token set significantly enhances biological language expressiveness, thereby improving the overall reasoning capacity of the model. Experimentally, our pretraining approach teaches protein models to self-correct and leads to substantial performance gains compared to standard pretraining.
Read the full abstract

All authors Xiang Zhang, Jiaqi Wei, Yuejin Yang, Zijie Qiu, Yuhan Chen, Zhiqiang Gao, Muhammad Abdul-Mageed, Laks V. S. Lakshmanan, Wanli Ouyang, Chenyu You, Siqi Sun

Computation and Language

MultiMind at SemEval-2025 Task 7: Crosslingual Fact-Checked Claim Retrieval via Multi-Source Alignment

Mohammad Mahdi Abootorabi, Alireza Ghahramani Kure, Mohammadali Mohammadkhani et al. (+2 more)

This paper presents our system for SemEval-2025 Task 7: Multilingual and Crosslingual Fact-Checked Claim Retrieval. In an era where misinformation spreads rapidly, effective fact-checking is increasingly critical. We introduce TriAligner, a novel approach that leverages a dual-encoder architecture with contrastive learning and incorporates both native and English translations across different modalities. Our method effectively retrieves claims across multiple languages by learning the relative importance of different sources in alignment. To enhance robustness, we employ efficient data preprocessing and augmentation using large language models while incorporating hard negative sampling to improve representation learning. We evaluate our approach on monolingual and crosslingual benchmarks, demonstrating significant improvements in retrieval accuracy and fact-checking performance over baselines.
Read the full abstract

All authors Mohammad Mahdi Abootorabi, Alireza Ghahramani Kure, Mohammadali Mohammadkhani, Sina Elahimanesh, Mohammad Ali Ali Panah

Computation and Language

Neural Probe-Based Hallucination Detection for Large Language Models

Shize Liang and Hongzhi Wang

Large language models(LLMs) excel at text generation and knowledge question-answering tasks, but they are prone to generating hallucinated content, severely limiting their application in high-risk domains. Current hallucination detection methods based on uncertainty estimation and external knowledge retrieval suffer from the limitation that they still produce erroneous content at high confidence levels and rely heavily on retrieval efficiency and knowledge coverage. In contrast, probe methods that leverage the model's hidden-layer states offer real-time and lightweight advantages. However, traditional linear probes struggle to capture nonlinear structures in deep semantic spaces.To overcome these limitations, we propose a neural network-based framework for token-level hallucination detection. By freezing language model parameters, we employ lightweight MLP probes to perform nonlinear modeling of high-level hidden states. A multi-objective joint loss function is designed to enhance detection stability and semantic disambiguity. Additionally, we establish a layer position-probe performance response model, using Bayesian optimization to automatically search for optimal probe insertion layers and achieve superior training results.Experimental results on LongFact, HealthBench, and TriviaQA demonstrate that MLP probes significantly outperform state-of-the-art methods in accuracy, recall, and detection capability under low false-positive conditions.
Read the full abstract

Computation and Language

Foundation Model-based Evaluation of Neuropsychiatric Disorders: A Lifespan-Inclusive, Multi-Modal, and Multi-Lingual Study

Zhongren Dong, Haotian Guo, Weixiang Xu et al. (+2 more)

Neuropsychiatric disorders, such as Alzheimer's disease (AD), depression, and autism spectrum disorder (ASD), are characterized by linguistic and acoustic abnormalities, offering potential biomarkers for early detection. Despite the promise of multi-modal approaches, challenges like multi-lingual generalization and the absence of a unified evaluation framework persist. To address these gaps, we propose FEND (Foundation model-based Evaluation of Neuropsychiatric Disorders), a comprehensive multi-modal framework integrating speech and text modalities for detecting AD, depression, and ASD across the lifespan. Leveraging 13 multi-lingual datasets spanning English, Chinese, Greek, French, and Dutch, we systematically evaluate multi-modal fusion performance. Our results show that multi-modal fusion excels in AD and depression detection but underperforms in ASD due to dataset heterogeneity. We also identify modality imbalance as a prevalent issue, where multi-modal fusion fails to surpass the best mono-modal models. Cross-corpus experiments reveal robust performance in task- and language-consistent scenarios but noticeable degradation in multi-lingual and task-heterogeneous settings. By providing extensive benchmarks and a detailed analysis of performance-influencing factors, FEND advances the field of automated, lifespan-inclusive, and multi-lingual neuropsychiatric disorder assessment. We encourage researchers to adopt the FEND framework for fair comparisons and reproducible research.
Read the full abstract

All authors Zhongren Dong, Haotian Guo, Weixiang Xu, Huan Zhao, Zixing Zhang

Computation and Language · Cross-listed from cs.CV

Transductive Visual Programming: Evolving Tool Libraries from Experience for Spatial Reasoning

Shengguang Wu, Xiaohan Wang, Yuhui Zhang et al. (+2 more)

Spatial reasoning in 3D scenes requires precise geometric calculations that challenge vision-language models. Visual programming addresses this by decomposing problems into steps calling specialized tools, yet existing methods rely on either fixed toolsets or speculative tool induction before solving problems, resulting in suboptimal programs and poor utilization of induced tools. We present Transductive Visual Programming (TVP), a novel framework that builds new tools from its own experience rather than speculation. TVP first solves problems using basic tools while accumulating experiential solutions into an Example Library, then abstracts recurring patterns from these programs into reusable higher-level tools for an evolving Tool Library. This allows TVP to tackle new problems with increasingly powerful tools learned from experience. On Omni3D-Bench, TVP achieves state-of-the-art performance, outperforming GPT-4o by 22% and the previous best visual programming system by 11%. Our transductively learned tools are used 5x more frequently as core program dependency than inductively created ones, demonstrating more effective tool discovery and reuse. The evolved tools also show strong generalization to unseen spatial tasks, achieving superior performance on benchmarks from SpatialScore-Hard collection without any testset-specific modification. Our work establishes experience-driven transductive tool creation as a powerful paradigm for building self-evolving visual programming agents that effectively tackle challenging spatial reasoning tasks. We release our code at https://transductive-visualprogram.github.io/.
Read the full abstract

All authors Shengguang Wu, Xiaohan Wang, Yuhui Zhang, Hao Zhu, Serena Yeung-Levy

Computation and Language · Cross-listed from q-bio.NC

Decoding Predictive Inference in Visual Language Processing via Spatiotemporal Neural Coherence

Sean C. Borneman, Julia Krebs, Ronnie B. Wilbur et al. (+1 more)

Human language processing relies on the brain's capacity for predictive inference. We present a machine learning framework for decoding neural (EEG) responses to dynamic visual language stimuli in Deaf signers. Using coherence between neural signals and optical flow-derived motion features, we construct spatiotemporal representations of predictive neural dynamics. Through entropy-based feature selection, we identify frequency-specific neural signatures that differentiate interpretable linguistic input from linguistically disrupted (time-reversed) stimuli. Our results reveal distributed left-hemispheric and frontal low-frequency coherence as key features in language comprehension, with experience-dependent neural signatures correlating with age. This work demonstrates a novel multimodal approach for probing experience-driven generative models of perception in the brain.
Read the full abstract

All authors Sean C. Borneman, Julia Krebs, Ronnie B. Wilbur, Evie A. Malaia

Computation and Language

Where Did This Sentence Come From? Tracing Provenance in LLM Reasoning Distillation

Kaiyuan Liu, Shaotian Yan, Rui Miao et al. (+4 more)

Reasoning distillation has attracted increasing attention. It typically leverages a large teacher model to generate reasoning paths, which are then used to fine-tune a student model so that it mimics the teacher's behavior in training contexts. However, previous approaches have lacked a detailed analysis of the origins of the distilled model's capabilities. It remains unclear whether the student can maintain consistent behaviors with the teacher in novel test-time contexts, or whether it regresses to its original output patterns, raising concerns about the generalization of distillation models. To analyse this question, we introduce a cross-model Reasoning Distillation Provenance Tracing framework. For each action (e.g., a sentence) produced by the distilled model, we obtain the predictive probabilities assigned by the teacher, the original student, and the distilled model under the same context. By comparing these probabilities, we classify each action into different categories. By systematically disentangling the provenance of each action, we experimentally demonstrate that, in test-time contexts, the distilled model can indeed generate teacher-originated actions, which correlate with and plausibly explain observed performance on distilled model. Building on this analysis, we further propose a teacher-guided data selection method. Unlike prior approach that rely on heuristics, our method directly compares teacher-student divergences on the training data, providing a principled selection criterion. We validate the effectiveness of our approach across multiple representative teacher models and diverse student models. The results highlight the utility of our provenance-tracing framework and underscore its promise for reasoning distillation. We hope to share Reasoning Distillation Provenance Tracing and our insights into reasoning distillation with the community.
Read the full abstract

All authors Kaiyuan Liu, Shaotian Yan, Rui Miao, Bing Wang, Chen Shen, Jun Zhang, Jieping Ye

Computation and Language

Architectural Trade-offs in Small Language Models Under Compute Constraints

Shivraj Singh Bhatti

We present a systematic empirical study of small language models under strict compute constraints, analyzing how architectural choices and training budget interact to determine performance. Starting from a linear next-token predictor, we progressively introduce nonlinearities, self-attention, and multi-layer transformer architectures, evaluating each on character-level modeling of Tiny Shakespeare and word-level modeling of Penn Treebank (PTB) and WikiText-2. We compare models using test negative log-likelihood (NLL), parameter count, and approximate training FLOPs to characterize accuracy-efficiency trade-offs. Our results show that attention-based models dominate MLPs in per-FLOP efficiency even at small scale, while increasing depth or context without sufficient optimization can degrade performance. We further examine rotary positional embeddings (RoPE), finding that architectural techniques successful in large language models do not necessarily transfer to small-model regimes.
Read the full abstract