Peer Revenue Impact factor: volatile

Papers

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

Computer Science Computation and Language

cs.CL Watch Computation and Language

Tuesday 21 October

Computation and Language

Grounding or Guessing? Visual Signals for Detecting Hallucinations in Sign Language Translation

Yasser Hamidullah, Koel Dutta Chowdhury, Yusser Al Ghussin et al. (+4 more)

Hallucination, where models generate fluent text unsupported by visual evidence, remains a major flaw in vision-language models and is particularly critical in sign language translation (SLT). In SLT, meaning depends on precise grounding in video, and gloss-free models are especially vulnerable because they map continuous signer movements directly into natural language without intermediate gloss supervision that serves as alignment. We argue that hallucinations arise when models rely on language priors rather than visual input. To capture this, we propose a token-level reliability measure that quantifies how much the decoder uses visual information. Our method combines feature-based sensitivity, which measures internal changes when video is masked, with counterfactual signals, which capture probability differences between clean and altered video inputs. These signals are aggregated into a sentence-level reliability score, providing a compact and interpretable measure of visual grounding. We evaluate the proposed measure on two SLT benchmarks (PHOENIX-2014T and CSL-Daily) with both gloss-based and gloss-free models. Our results show that reliability predicts hallucination rates, generalizes across datasets and architectures, and decreases under visual degradations. Beyond these quantitative trends, we also find that reliability distinguishes grounded tokens from guessed ones, allowing risk estimation without references; when combined with text-based signals (confidence, perplexity, or entropy), it further improves hallucination risk estimation. Qualitative analysis highlights why gloss-free models are more susceptible to hallucinations. Taken together, our findings establish reliability as a practical and reusable tool for diagnosing hallucinations in SLT, and lay the groundwork for more robust hallucination detection in multimodal generation.
Read the full abstract

All authors Yasser Hamidullah, Koel Dutta Chowdhury, Yusser Al Ghussin, Shakib Yazdani, Cennet Oguz, Josef van Genabith, Cristina España-Bonet

Computation and Language

Chain-of-Conceptual-Thought Elicits Daily Conversation in Large Language Models

Qingqing Gu, Dan Wang, Yue Zhao et al. (+5 more)

Chain-of-Thought (CoT) is widely applied to enhance the LLM capability in math, coding and reasoning tasks. However, its performance is limited for open-domain tasks, when there are no clearly defined reasoning steps or logical transitions. To mitigate such challenges, we propose a new prompt-based paradigm called Chain of Conceptual Thoughts (CoCT), which suggests the LLM first to produce the tag of concepts, then complete the detailed content following the concept. To encourage this hierarchical way of thinking, we implement the concepts with emotions, strategies and topics. We experiment with this paradigm in daily and emotional support conversations, covering tasks with both in-domain and out-of-domain concept settings. Automatic, human, and LLM-based evaluations reveal that CoCT surpasses several prompt-based baselines such as self-refine, ECoT, SoT and RAG, suggesting a potential solution of LLM prompting paradigm for a wider scope of tasks.
Read the full abstract

All authors Qingqing Gu, Dan Wang, Yue Zhao, Xiaoyu Wang, Zhonglin Jiang, Yong Chen, Hongyan Li, Luo Ji

Computation and Language

Adamas: Hadamard Sparse Attention for Efficient Long-Context Inference

Siyuan Yan, Guo-Qing Jiang, Yuchen Zhang et al. (+4 more)

Large language models (LLMs) now support context windows of hundreds of thousands to millions of tokens, enabling applications such as long-document summarization, large-scale code synthesis, multi-document question answering and persistent multi-turn dialogue. However, such extended contexts exacerbate the quadratic cost of self-attention, leading to severe latency in autoregressive decoding. Existing sparse attention methods reduce these costs by attending to only a subset of key-value (KV) tokens, but fixed or coarse-grained selection strategies can miss important tokens or retain substantial redundancy under constrained token budgets. We introduce Adamas, an accurate and efficient token-level sparse attention mechanism for long-context inference. Adamas applies the Hadamard transform, bucketization and 2-bit compression to produce compact representations, and uses lightweight distance-based estimation for efficient top-$k$ candidate selection. The selected positions are then used to gather the corresponding original KV states for sparse attention. Comprehensive evaluations show that Adamas maintains near-lossless performance within the effective context length of Llama-3.1-8B-Instruct. In efficiency evaluations, Adamas achieves up to $6.5\times$ self-attention speedup and $1.7\times$ end-to-end decoding speedup at a 32K context length. Adamas also maintains performance comparable to full attention on challenging long-form reasoning benchmarks, including AIME 2024 and AIME 2025.
Read the full abstract

All authors Siyuan Yan, Guo-Qing Jiang, Yuchen Zhang, Xiaoxing Ma, Ran Zhu, Chun Cao, Jingwei Xu

Computation and Language

MENTOR: Reinforcement Learning via Flexible Teacher-Optimized Rewards for Tool-Use Distillation

ChangSu Choi, Hoyun Song, Dongyeon Kim et al. (+6 more)

Distilling the tool-use capabilities of large language models (LLMs) into small language models (SLMs) is essential for their practical application. The predominant approach, supervised fine-tuning (SFT), is an off-policy distillation method that suffers from poor out-of-domain (OOD) generalization because it rigidly aligns with static teacher trajectories. While reinforcement learning (RL) offers an alternative, the capacity limitations of SLMs pose a severe dilemma: sparse outcome rewards provide insufficient guidance, whereas strict trajectory matching imposes overly restrictive constraints. To bridge this capacity-driven gap, we propose MENTOR, an on-policy distillation framework that introduces a flexible yet process-aware reward structure. Instead of enforcing rigid replication, MENTOR uses the teacher's reference to guide tool-use behavior, balancing behavioral alignment with downstream performance. Extensive experiments on controlled executable-tool benchmarks demonstrate that MENTOR improves OOD tool-use performance compared to SFT and strict RL baselines. Our findings suggest that within verifiable tool-use environments, flexible tool-use alignment offers a more effective approach than strict trajectory replication for developing adaptable small models.
Read the full abstract

All authors ChangSu Choi, Hoyun Song, Dongyeon Kim, Minkyung Cho, WooHyeon Jung, Sunjin Park, NohHyeob Bae, Seona Yu, KyungTae Lim

Computation and Language

Towards Fair ASR For Second Language Speakers Using Fairness Prompted Finetuning

Monorama Swain, Bubai Maji, Jagabandhu Mishra et al. (+3 more)

In this work, we address the challenge of building fair English ASR systems for second-language speakers. Our analysis of widely used ASR models, Whisper and Seamless-M4T, reveals large fluctuations in word error rate (WER) across 26 accent groups, indicating significant fairness gaps. To mitigate this, we propose fairness-prompted finetuning with lightweight adapters, incorporating Spectral Decoupling (SD), Group Distributionally Robust Optimization (Group-DRO), and Invariant Risk Minimization (IRM). Our proposed fusion of traditional empirical risk minimization (ERM) with cross-entropy and fairness-driven objectives (SD, Group DRO, and IRM) enhances fairness across accent groups while maintaining overall recognition accuracy. In terms of macro-averaged word error rate, our approach achieves a relative improvement of 58.7% and 58.5% over the large pretrained Whisper and SeamlessM4T, and 9.7% and 7.8% over them, finetuning with standard empirical risk minimization with cross-entropy loss.
Read the full abstract

All authors Monorama Swain, Bubai Maji, Jagabandhu Mishra, Markus Schedl, Anders Søgaard, Jesper Rindom Jensen

Computation and Language

KoSimpleQA: A Korean Factuality Benchmark with an Analysis of Reasoning LLMs

Donghyeon Ko, Kyubyung Chae, Yeguk Jin et al. (+6 more)

We present $\textbf{Korean SimpleQA (KoSimpleQA)}$, a benchmark for evaluating factuality in large language models (LLMs) with a focus on Korean cultural knowledge. KoSimpleQA is designed to be challenging yet easy to grade, consisting of 938 short, fact-seeking questions with unambiguous answers. We conduct a comprehensive evaluation across a diverse set of open-source LLMs of varying sizes that support Korean, and find that even the strongest model generates correct answer only 31.6% of the time, underscoring the challenging nature of KoSimpleQA. Notably, performance rankings on KoSimpleQA differ substantially from those on the English SimpleQA, highlighting the unique value of our dataset. Furthermore, we observe that reasoning helps mitigate the cross-lingual knowledge gap in LLMs, which refers to disparities in their ability to manifest knowledge across languages. KoSimpleQA can be found at https://github.com/naver-ai/KoSimpleQA.
Read the full abstract

All authors Donghyeon Ko, Kyubyung Chae, Yeguk Jin, Byungwook Lee, Chansong Jo, Sookyo In, Jaehong Lee, Taesup Kim, Donghyun Kwak

Computation and Language

KrishokBondhu: A Retrieval-Augmented Voice-Based Agricultural Advisory Call Center for Bengali Farmers

Mohd Ruhul Ameen, Akif Islam, Farjana Aktar et al. (+1 more)

In Bangladesh, many farmers still struggle to access timely, expert-level agricultural guidance. This paper presents KrishokBondhu, a voice-enabled, call-centre-integrated advisory platform built on a Retrieval-Augmented Generation (RAG) framework for Bengali-speaking farmers. The system combines agricultural handbooks, extension manuals, and NGO publications, processes them through an OCR-based pipeline, and indexes the curated content in a vector database for semantic retrieval. Through a phone-based interface, farmers can receive real-time, context-aware advice: speech-to-text converts the Bengali query, the RAG module retrieves relevant information, a large language model (Gemma 3-4B) generates a grounded response, and text-to-speech delivers the answer in spoken Bengali. In a pilot evaluation, KrishokBondhu produced high-quality responses for 72.7% of diverse agricultural queries. Compared to the KisanQRS benchmark, it achieved a composite score of 4.53 versus 3.13 on a 5-point scale, with a 44.7% improvement and especially large gains in contextual richness and completeness, while maintaining comparable relevance and technical specificity. Semantic-similarity analysis further showed a strong correlation between retrieved context and answer quality. KrishokBondhu demonstrates the feasibility of combining call-centre accessibility, multilingual voice interaction, and modern RAG techniques to deliver expert-level agricultural guidance to remote Bangladeshi farmers.
Read the full abstract

All authors Mohd Ruhul Ameen, Akif Islam, Farjana Aktar, M. Saifuzzaman Rafat

Computation and Language

Combining Distantly Supervised Models with In Context Learning for Monolingual and Cross-Lingual Relation Extraction

Vipul Rathore, Malik Hammad Faisal, Parag Singla et al. (+1 more)

Distantly Supervised Relation Extraction (DSRE) remains a long-standing challenge in NLP, where models must learn from noisy bag-level annotations while making sentence-level predictions. While existing state-of-the-art (SoTA) DSRE models rely on task-specific training, their integration with in-context learning (ICL) using large language models (LLMs) remains underexplored. A key challenge is that the LLM may not learn relation semantics correctly, due to noisy annotation. In response, we propose HYDRE – HYbrid Distantly Supervised Relation Extraction framework. It first uses a trained DSRE model to identify the top-k candidate relations for a given test sentence, then uses a novel dynamic exemplar retrieval strategy that extracts reliable, sentence-level exemplars from training data, which are then provided in LLM prompt for outputting the final relation(s). We further extend HYDRE to cross-lingual settings for RE in low-resource languages. Using available English DSRE training data, we evaluate all methods on English as well as a newly curated benchmark covering four diverse low-resource Indic languages – Oriya, Santali, Manipuri, and Tulu. HYDRE achieves up to 20 F1 point gains in English and, on average, 17 F1 points on Indic languages over prior SoTA DSRE models. Detailed ablations exhibit HYDRE's efficacy compared to other prompting strategies.
Read the full abstract

All authors Vipul Rathore, Malik Hammad Faisal, Parag Singla, Mausam

Computation and Language

ECG-LLM – training and evaluation of domain-specific large language models for electrocardiography

Lara Ahrens, Wilhelm Haverkamp and Nils Strodthoff

Domain-adapted open-weight large language models (LLMs) offer promising healthcare applications, from queryable knowledge bases to multimodal assistants, with the crucial advantage of local deployment for privacy preservation. However, optimal adaptation strategies, evaluation methodologies, and performance relative to general-purpose LLMs remain poorly characterized. We investigated these questions in electrocardiography, an important area of cardiovascular medicine, by finetuning open-weight models on domain-specific literature and implementing a multi-layered evaluation framework comparing finetuned models, retrieval-augmented generation (RAG), and Claude Sonnet 3.7 as a representative general-purpose model. Finetuned Llama 3.1 70B achieved superior performance on multiple-choice evaluations and automatic text metrics, ranking second to Claude 3.7 in LLM-as-a-judge assessments. Human expert evaluation favored Claude 3.7 and RAG approaches for complex queries. Finetuned models significantly outperformed their base counterparts across nearly all evaluation modes. Our findings reveal substantial performance heterogeneity across evaluation methodologies, underscoring assessment complexity. Nevertheless, domain-specific adaptation through finetuning and RAG achieves competitive performance with proprietary models, supporting the viability of privacy-preserving, locally deployable clinical solutions.
Read the full abstract

Computation and Language · Cross-listed from cs.CR

Position: LLM Watermarking Should Align Stakeholders' Incentives for Practical Adoption

Yepeng Liu, Xuandong Zhao, Dawn Song et al. (+2 more)

Despite progress in watermarking algorithms for large language models (LLMs), real-world deployment remains limited. We argue that this gap stems from misaligned incentives among LLM providers, platforms, and end users, which manifest as three key barriers: competitive risk, detection-tool governance, and attribution issues. We revisit three classes of watermarking through this lens. Model watermarking naturally aligns with LLM provider interests, yet faces new challenges in open-source ecosystems. LLM text watermarking offers modest provider benefit when framed solely as an anti-misuse tool, but can gain traction in narrowly scoped settings such as dataset de-contamination or user-controlled provenance. In-context watermarking (ICW) is tailored for trusted parties, such as conference organizers or educators, who embed hidden watermarking instructions into documents. If a dishonest reviewer or student submits this text to an LLM, the output carries a detectable watermark indicating misuse. This setup aligns incentives: users experience no quality loss, trusted parties gain a detection tool, and LLM providers remain neutral by simply following watermark instructions. We advocate for a broader exploration of incentive-aligned methods, with ICW as an example, in domains where trusted parties need reliable tools to detect misuse. More broadly, we distill design principles for incentive-aligned, domain-specific watermarking and outline future research directions. Our position is that the practical adoption of LLM watermarking requires aligning stakeholder incentives in targeted application domains and fostering active community engagement.
Read the full abstract

All authors Yepeng Liu, Xuandong Zhao, Dawn Song, Gregory W. Wornell, Yuheng Bu

Computation and Language · Cross-listed from cs.CV

The Impact of Image Resolution on Biomedical Multimodal Large Language Models

Liangyu Chen, James Burgess, Jeffrey J Nirschl et al. (+2 more)

Imaging technologies are fundamental to biomedical research and modern medicine, requiring analysis of high-resolution images across various modalities. While multimodal large language models (MLLMs) show promise for biomedical image analysis, most are designed for low-resolution images from general-purpose datasets, risking critical information loss. We investigate how image resolution affects MLLM performance in biomedical applications and demonstrate that: (1) native-resolution training and inference significantly improve performance across multiple tasks, (2) misalignment between training and inference resolutions severely degrades performance, and (3) mixed-resolution training effectively mitigates misalignment and balances computational constraints with performance requirements. Based on these findings, we recommend prioritizing native-resolution inference and mixed-resolution datasets to optimize biomedical MLLMs for transformative impact in scientific research and clinical applications.
Read the full abstract

All authors Liangyu Chen, James Burgess, Jeffrey J Nirschl, Orr Zohar, Serena Yeung-Levy

Computation and Language

From Retrieval to Generation: Unifying External and Parametric Knowledge for Medical Question Answering

Lei Li, Xiao Zhou, Yingying Zhang et al. (+1 more)

Medical question answering (QA) requires extensive access to domain-specific knowledge. A promising direction is to enhance large language models (LLMs) with external knowledge retrieved from medical corpora or parametric knowledge stored in model parameters. Existing approaches typically fall into two categories: Retrieval-Augmented Generation (RAG), which grounds model reasoning on externally retrieved evidence, and Generation-Augmented Generation (GAG), which depends solely on the models internal knowledge to generate contextual documents. However, RAG often suffers from noisy or incomplete retrieval, while GAG is vulnerable to hallucinated or inaccurate information due to unconstrained generation. Both issues can mislead reasoning and undermine answer reliability. To address these challenges, we propose MedRGAG, a unified retrieval-generation augmented framework that seamlessly integrates external and parametric knowledge for medical QA. MedRGAG comprises two key modules: Knowledge-Guided Context Completion (KGCC), which directs the generator to produce background documents that complement the missing knowledge revealed by retrieval; and Knowledge-Aware Document Selection (KADS), which adaptively selects an optimal combination of retrieved and generated documents to form concise yet comprehensive evidence for answer generation. Extensive experiments on five medical QA benchmarks demonstrate that MedRGAG achieves a 12.5% improvement over MedRAG and a 4.5% gain over MedGENIE, highlighting the effectiveness of unifying retrieval and generation for knowledge-intensive reasoning. Our code and data are publicly available at https://anonymous.4open.science/r/MedRGAG
Read the full abstract

All authors Lei Li, Xiao Zhou, Yingying Zhang, Xian Wu

Computation and Language

Food4All: An Agentic Framework and Benchmark for Food Resource Navigation with Adaptive User Understanding

Yiyang Li, Weixiang Sun, Tianyi Ma et al. (+3 more)

Food assistance referral requires conversational agents to translate underspecified, often noisy help-seeking dialogues into locally valid resource recommendations. We present Food4All, an agentic food-resource referral framework and benchmark grounded in 686 structured Indiana food resources. Food4All couples a food-specific search tool with 300 multi-turn evaluation tasks spanning single food needs, composite cases with access or document constraints, and five non-ideal user interaction traits: unreasonable demands, rambling responses, impatience, incomplete answers, and inconsistent information. We evaluate six Large Language Models (LLMs) on requirement grounding, resource retrieval, final referral correctness, and interaction efficiency. Although the strongest model achieves 96.33% referral accuracy, our diagnostics reveal persistent failures in grounding schedule, eligibility, intake, and document constraints, as well as failures to preserve valid retrieved resources in the final recommendation. Trait-level analysis further shows that different non-ideal behaviors stress different parts of the referral pipeline. Food4All provides a controlled testbed for studying tool-calling agents in constraint-sensitive food assistance referral under realistic user interaction challenges.
Read the full abstract

All authors Yiyang Li, Weixiang Sun, Tianyi Ma, Kaiwen Shi, Zheyuan Zhang, Yanfang Ye

2510.18289 · 21 Oct · We have further refined the benchmark construction and experimental presentation to improve clarity and consistency. The revised version includes updated task design, food-resource data, and evaluation details to better align the benchmark with the intended food resource referral setting. These changes provide a more precise presentation of the experimental findings

Abstract PDF $AGNT ×2$BNCH ×2$LANG ×1$NAVIG ×1 +1

Computation and Language

BrailleLLM: Braille Instruction Tuning with Large Language Models for Braille Domain Tasks

Tianyuan Huang, Zepeng Zhu, Hangdi Xing et al. (+6 more)

Braille plays a vital role in education and information accessibility for visually impaired individuals. However, Braille information processing faces challenges such as data scarcity and ambiguities in mixed-text contexts. We construct English and Chinese Braille Mixed Datasets (EBMD/CBMD) with mathematical formulas to support diverse Braille domain research, and propose a syntax tree-based augmentation method tailored for Braille data. To address the underperformance of traditional fine-tuning methods in Braille-related tasks, we investigate Braille Knowledge-Based Fine-Tuning (BKFT), which reduces the learning difficulty of Braille contextual features. BrailleLLM employs BKFT via instruction tuning to achieve unified Braille translation, formula-to-Braille conversion, and mixed-text translation. Experiments demonstrate that BKFT achieves significant performance improvements over conventional fine-tuning in Braille translation scenarios. Our open-sourced datasets and methodologies establish a foundation for low-resource multilingual Braille research.
Read the full abstract

All authors Tianyuan Huang, Zepeng Zhu, Hangdi Xing, Zirui Shao, Zhi Yu, Chaoxiong Yang, Jiaxian He, Xiaozhong Liu, Jiajun Bu

Computation and Language

Text or Pixels? It Takes Half: On the Token Efficiency of Visual Text Inputs in Multimodal LLMs

Yanhong Li, Zixuan Lan and Jiawei Zhou

Large language models (LLMs) and their multimodal variants can now process visual inputs, including images of text. This raises an intriguing question: can we compress textual inputs by feeding them as images to reduce token usage while preserving performance? In this paper, we show that visual text representations are a practical and surprisingly effective form of input compression for decoder LLMs. We exploit the idea of rendering long text inputs as a single image and provide it directly to the model. This leads to dramatically reduced number of decoder tokens required, offering a new form of input compression. Through experiments on two distinct benchmarks RULER (long-context retrieval) and CNN/DailyMail (document summarization) we demonstrate that this text-as-image method yields substantial token savings (often nearly half) without degrading task performance.
Read the full abstract

Computation and Language

DelvePO: Direction-Guided Self-Evolving Framework for Flexible Prompt Optimization

Tao Tao, Guanghui Zhu, Lang Guo et al. (+3 more)

Prompt Optimization has emerged as a crucial approach due to its capabilities in steering Large Language Models to solve various tasks. However, current works mainly rely on the random rewriting ability of LLMs, and the optimization process generally focus on specific influencing factors, which makes it easy to fall into local optimum. Besides, the performance of the optimized prompt is often unstable, which limits its transferability in different tasks. To address the above challenges, we propose $\textbf{DelvePO}$ ($\textbf{D}$irection-Guid$\textbf{e}$d Se$\textbf{l}$f-E$\textbf{v}$olving Framework for Fl$\textbf{e}$xible $\textbf{P}$rompt $\textbf{O}$ptimization), a task-agnostic framework to optimize prompts in self-evolve manner. In our framework, we decouple prompts into different components that can be used to explore the impact that different factors may have on various tasks. On this basis, we introduce working memory, through which LLMs can alleviate the deficiencies caused by their own uncertainties and further obtain key insights to guide the generation of new prompts. Extensive experiments conducted on different tasks covering various domains for both open- and closed-source LLMs, including DeepSeek-R1-Distill-Llama-8B, Qwen2.5-7B-Instruct and GPT-4o-mini. Experimental results show that DelvePO consistently outperforms previous SOTA methods under identical experimental settings, demonstrating its effectiveness and transferability across different tasks.
Read the full abstract

All authors Tao Tao, Guanghui Zhu, Lang Guo, Hongyi Chen, Chunfeng Yuan, Yihua Huang

Computation and Language · Cross-listed from cs.CV

VLSU: Mapping the Limits of Joint Multimodal Understanding for AI Safety

Shruti Palaskar, Leon Gatys, Mona Abdelrahman et al. (+9 more)

Safety evaluation of multimodal foundation models often treats vision and language inputs separately, missing risks from joint interpretation where benign content becomes harmful in combination. Existing approaches also fail to distinguish clearly unsafe content from borderline cases, leading to problematic over-blocking or under-refusal of genuinely harmful content. We present Vision Language Safety Understanding (VLSU), a comprehensive framework to systematically evaluate multimodal safety through fine-grained severity classification and combinatorial analysis across 17 distinct safety patterns. Using a multi-stage pipeline with real-world images and human annotation, we construct a large-scale benchmark of 8,187 samples spanning 15 harm categories. Our evaluation of eleven state-of-the-art models reveals systematic joint understanding failures: while models achieve 90%-plus accuracy on clear unimodal safety signals, performance degrades substantially to 20-55% when joint image-text reasoning is required to determine the safety label. Most critically, 34% of errors in joint image-text safety classification occur despite correct classification of the individual modalities, further demonstrating absent compositional reasoning capabilities. Additionally, we find that models struggle to balance refusing unsafe content while still responding to borderline cases that deserve engagement. For example, we find that instruction framing can reduce the over-blocking rate on borderline content from 62.4% to 10.4% in Gemini-1.5, but only at the cost of under-refusing on unsafe content with refusal rate dropping from 90.8% to 53.9%. Overall, our framework exposes weaknesses in joint image-text understanding and alignment gaps in current models, and provides a critical test bed to enable the next milestones in research on robust vision-language safety.
Read the full abstract

All authors Shruti Palaskar, Leon Gatys, Mona Abdelrahman, Mar Jacobo, Larry Lindsey, Rutika Moharir, Gunnar Lund, Yang Xu, Navid Shiee, Jeffrey Bigham, Charles Maalouf, Joseph Yitan Cheng

Computation and Language

MARCUS: An Event-Centric NLP Pipeline that generates Character Arcs from Narratives

Sriharsh Bhyravajjula, Ujwal Narayan and Manish Shrivastava

Character arcs are important theoretical devices employed in literary studies to understand character journeys, identify tropes across literary genres, and establish similarities between narratives. This work addresses the novel task of computationally generating event-centric, relation-based character arcs from narratives. Providing a quantitative representation for arcs brings tangibility to a theoretical concept and paves the way for subsequent applications. We present MARCUS (Modelling Arcs for Understanding Stories), an NLP pipeline that extracts events, participant characters, implied emotion, and sentiment to model inter-character relations. MARCUS tracks and aggregates these relations across the narrative to generate character arcs as graphical plots. We generate character arcs from two extended fantasy series, Harry Potter and Lord of the Rings. We evaluate our approach before outlining existing challenges, suggesting applications of our pipeline, and discussing future work.
Read the full abstract

Computation and Language

Contrastive Decoding Mitigates Score Range Bias in LLM-as-a-Judge

Yoshinari Fujinuma

Large Language Models (LLMs) are commonly used as evaluators in various applications, but the reliability of the outcomes remains a challenge. One such challenge is using LLMs-as-judges for direct assessment, i.e., assigning scores from a specified range without any references. Focusing on summarization, we first show that this challenge stems from LLM judge outputs being associated with score range bias, i.e., LLM judge outputs are highly sensitive to pre-defined score ranges. We also show that similar biases exist among models from the same family. We then mitigate this bias through contrastive decoding, achieving up to 11.7% relative improvement on average in Spearman correlation with human judgments across different score ranges.
Read the full abstract

Monday 20 October

Computation and Language · Cross-listed from cs.AI

Naiad: Novel Agentic Intelligent Autonomous System for Inland Water Monitoring

Eirini Baltzi, Tilemachos Moumouris, Athena Psalta et al. (+2 more)

Inland water monitoring is vital for safeguarding public health and ecosystems, enabling timely interventions to mitigate risks. Existing methods often address isolated sub-problems such as cyanobacteria, chlorophyll, or other quality indicators separately. NAIAD introduces an agentic AI assistant that leverages Large Language Models (LLMs) and external analytical tools to deliver a holistic solution for inland water monitoring using Earth Observation (EO) data. Designed for both experts and non-experts, NAIAD provides a single-prompt interface that translates natural-language queries into actionable insights. Through Retrieval-Augmented Generation (RAG), LLM reasoning, external tool orchestration, computational graph execution, and agentic reflection, it retrieves and synthesizes knowledge from curated sources to produce tailored reports. The system integrates diverse tools for weather data, Sentinel-2 imagery, remote-sensing index computation (e.g., NDCI), chlorophyll-a estimation, and established platforms such as CyFi. Performance is evaluated using correctness and relevancy metrics, achieving over 77% and 85% respectively on a dedicated benchmark covering multiple user-expertise levels. Preliminary results show strong adaptability and robustness across query types. An ablation study on LLM backbones further highlights Gemma 3 (27B) and Qwen 2.5 (14B) as offering the best balance between computational efficiency and reasoning performance.
Read the full abstract

All authors Eirini Baltzi, Tilemachos Moumouris, Athena Psalta, Vasileios Tsironis, Konstantinos Karantzalos

Computation and Language · Cross-listed from cs.DC

Network and Systems Performance Characterization of MCP-Enabled LLM Agents

Zihao Ding, Mufeng Zhu and Yao Liu

Model Context Protocol (MCP) has recently gained increased attention within the AI community for providing a standardized way for large language models (LLMs) to interact with external tools and services, significantly enhancing their capabilities. However, the inclusion of extensive contextual information, including system prompts, MCP tool definitions, and context histories, in MCP-enabled LLM interactions, dramatically inflates token usage. Given that LLM providers charge based on tokens, these expanded contexts can quickly escalate monetary costs and increase the computational load on LLM services. This paper presents a comprehensive measurement-based analysis of MCP-enabled interactions with LLMs, revealing trade-offs between capability, performance, and cost. We explore how different LLM models and MCP configurations impact key performance metrics such as token efficiency, monetary cost, task completion times, and task success rates, and suggest potential optimizations, including enabling parallel tool calls and implementing robust task abort mechanisms. These findings provide useful insights for developing more efficient, robust, and cost-effective MCP-enabled workflows.
Read the full abstract

Computation and Language · Cross-listed from cs.LG

An Enhanced Dual Transformer Contrastive Network for Multimodal Sentiment Analysis

Phuong Q. Dao, Mark Roantree and Vuong M. Ngo

Multimodal Sentiment Analysis (MSA) seeks to understand human emotions by jointly analyzing data from multiple modalities typically text and images offering a richer and more accurate interpretation than unimodal approaches. In this paper, we first propose BERT-ViT-EF, a novel model that combines powerful Transformer-based encoders BERT for textual input and ViT for visual input through an early fusion strategy. This approach facilitates deeper cross-modal interactions and more effective joint representation learning. To further enhance the model's capability, we propose an extension called the Dual Transformer Contrastive Network (DTCN), which builds upon BERT-ViT-EF. DTCN incorporates an additional Transformer encoder layer after BERT to refine textual context (before fusion) and employs contrastive learning to align text and image representations, fostering robust multimodal feature learning. Empirical results on two widely used MSA benchmarks MVSA-Single and TumEmo demonstrate the effectiveness of our approach. DTCN achieves best accuracy (78.4%) and F1-score (78.3%) on TumEmo, and delivers competitive performance on MVSA-Single, with 76.6% accuracy and 75.9% F1-score. These improvements highlight the benefits of early fusion and deeper contextual modeling in Transformer-based multimodal sentiment analysis.
Read the full abstract

Computation and Language

Transformer-Based Low-Resource Language Translation: A Study on Standard Bengali to Sylheti

Mangsura Kabir Oni and Tabia Tanzin Prama

Machine Translation (MT) has advanced from rule-based and statistical methods to neural approaches based on the Transformer architecture. While these methods have achieved impressive results for high-resource languages, low-resource varieties such as Sylheti remain underexplored. In this work, we investigate Bengali-to-Sylheti translation by fine-tuning multilingual Transformer models and comparing them with zero-shot large language models (LLMs). Experimental results demonstrate that fine-tuned models significantly outperform LLMs, with mBART-50 achieving the highest translation adequacy and MarianMT showing the strongest character-level fidelity. These findings highlight the importance of task-specific adaptation for underrepresented languages and contribute to ongoing efforts toward inclusive language technologies.
Read the full abstract

Computation and Language

Moneyball with LLMs: Analyzing Tabular Summarization in Sports Narratives

Ritam Upadhyay, Naman Ahuja, Rishabh Baral et al. (+2 more)

Large language model (LLM) approaches to tabular summarization rely on extensive prompt engineering, decomposition pipelines, or entity-level intermediate representations to achieve strong performance. While effective, these strategies are computationally expensive and offer limited insight into how well models maintain state over long, evolving narratives. We introduce SPORTABSET, a diagnostic benchmark for long-context tabular summarization across two complementary sports domains that require tracking multiple entities and aggregating statistics under domain-specific rules. Using SporTabSet, we systematically evaluate decomposition-based strategies across several long context LLMs. Results show that although decomposition substantially improves accuracy and numerical fidelity, gains stem mainly from dissecting multi-entity interference rather than improved local arithmetic. Robustness experiments further reveal high sensitivity to surface-level cues with structured failures, including hallucination, omission, and role confusion. Together, these findings identify consistent multientity memory as a key bottleneck in long context table generation, motivating diagnostic evaluation as a prerequisite for scalable, efficient and reliable tabular summarization models.
Read the full abstract

All authors Ritam Upadhyay, Naman Ahuja, Rishabh Baral, Aparna Garimella, Vivek Gupta

Computation and Language · Cross-listed from cs.AI

Saber: An Efficient Sampling with Adaptive Acceleration and Backtracking Enhanced Remasking for Diffusion Language Model

Yihong Dong, Zhaoyu Ma, Xue Jiang et al. (+10 more)

Diffusion language models (DLMs) are emerging as a compelling alternative to the dominant autoregressive paradigm, offering inherent advantages in parallel generation and bidirectional context modeling. However, for the tasks with strict structural constraints such as code generation, DLMs face a critical trade-off between inference speed and output quality, where accelerating generation by reducing sampling steps often leads to catastrophic performance collapse. We find that the fundamental reasons are: 1) the generation difficulty is non-uniform in the structured sequence decoding steps, making DLM's static acceleration strategy suboptimal; 2) the context of tokens generated by DLM evolves continuously, causing early high-confidence predictions to turn into irreversible errors. In this paper, we introduce efficient Sampling with Adaptive acceleration and Backtracking Enhanced Remasking (i.e., Saber), a novel training-free sampling algorithm for DLMs that first achieves both better inference speed and output quality in code generation. Saber dynamically adjusts the number of tokens unmasked per step based on the model's evolving confidence, and utilizes a backtracking mechanism to revert tokens whose confidence drops as new context emerges, with its effectiveness supported by theoretical analysis. Extensive experiments on multiple mainstream code generation benchmarks show that Saber boosts Pass@1 accuracy by an average of 1.9% over mainstream DLM sampling methods, while achieving an average 251.4% inference speedup. By leveraging the inherent advantages of DLMs, our work significantly narrows the performance gap with autoregressive models in code generation.
Read the full abstract

All authors Yihong Dong, Zhaoyu Ma, Xue Jiang, Zhiyuan Fan, Jiaru Qian, Yongmin Li, Jianha Xiao, Zhi Jin, Rongyu Cao, Binhua Li, Fei Huang, Yongbin Li, Ge Li

Computation and Language

Automatic Prompt Generation via Adaptive Selection of Prompting Techniques

Yohei Ikenoue, Hitomi Tashiro and Shigeru Kuroyanagi

Prompt engineering is crucial for achieving reliable and effective outputs from large language models (LLMs), but its design requires specialized knowledge of prompting techniques and a deep understanding of target tasks. To address this challenge, we propose a novel method that adaptively selects task-appropriate prompting techniques based on users' abstract task descriptions and automatically generates high-quality prompts without relying on pre-existing templates or frameworks. The proposed method constructs a knowledge base that associates task clusters, characterized by semantic similarity across diverse tasks, with their corresponding prompting techniques. When users input task descriptions, the system assigns them to the most relevant task cluster and dynamically generates prompts by integrating techniques drawn from the knowledge base. An experimental evaluation of the proposed method on 23 tasks from BIG-Bench Extra Hard (BBEH) demonstrates superior performance compared with standard prompts and existing automatic prompt-generation tools, as measured by both arithmetic and harmonic mean scores. This research establishes a foundation for streamlining and standardizing prompt creation, enabling non-experts to effectively leverage LLMs.
Read the full abstract

Computation and Language

Extracting Rule-based Descriptions of Attention Features in Transformers

Dan Friedman, Adithya Bhaskar, Alexander Wettig et al. (+1 more)

Mechanistic interpretability strives to explain model behavior in terms of bottom-up primitives. The leading paradigm is to express hidden states as a sparse linear combination of basis vectors, called features. However, this only identifies which text sequences (exemplars) activate which features; the actual interpretation of features requires subjective inspection of these exemplars. This paper advocates for a different solution: rule-based descriptions that match token patterns in the input and correspondingly increase or decrease the likelihood of specific output tokens. Specifically, we extract rule-based descriptions of SAE features trained on the outputs of attention layers. While prior work treats the attention layers as an opaque box, we describe how it may naturally be expressed in terms of interactions between input and output features, of which we study three types: (1) skip-gram rules of the form "[Canadian city]... speaks –> English", (2) absence rules of the form "[Montreal]... speaks -/-> English," and (3) counting rules that toggle only when the count of a word exceeds a certain value or the count of another word. Absence and counting rules are not readily discovered by inspection of exemplars, where manual and automatic descriptions often identify misleading or incomplete explanations. We then describe a simple approach to extract these types of rules automatically from a transformer, and apply it to GPT-2 small. We find that a majority of features may be described well with around 100 skip-gram rules, though absence rules are abundant even as early as the first layer (in over a fourth of features). We also isolate a few examples of counting rules. This paper lays the groundwork for future research into rule-based descriptions of features by defining them, showing how they may be extracted, and providing a preliminary taxonomy of some of the behaviors they represent.
Read the full abstract

All authors Dan Friedman, Adithya Bhaskar, Alexander Wettig, Danqi Chen

Computation and Language

LLMs Encode How Difficult Problems Are

William Lugoloobi and Chris Russell

Large language models exhibit a puzzling inconsistency: they solve complex problems yet frequently fail on seemingly simpler ones. We investigate whether LLMs internally encode problem difficulty in a way that aligns with human judgment, and whether this representation tracks generalization during reinforcement learning post-training. We train linear probes across layers and token positions on 60 models, evaluating on mathematical and coding subsets of Easy2HardBench. We find that human-labeled difficulty is strongly linearly decodable (AMC: $\rho \approx 0.88$) and exhibits clear model-size scaling, whereas LLM-derived difficulty is substantially weaker and scales poorly. Steering along the difficulty direction reveals that pushing models toward "easier" representations reduces hallucination and improves accuracy. During GRPO training on Qwen2.5-Math-1.5B, the human-difficulty probe strengthens and positively correlates with test accuracy across training steps, while the LLM-difficulty probe degrades and negatively correlates with performance. These results suggest that human annotations provide a stable difficulty signal that RL amplifies, while automated difficulty estimates derived from model performance become misaligned precisely as models improve. We release probe code and evaluation scripts to facilitate replication.
Read the full abstract

Computation and Language · Cross-listed from cs.CV

SafeCoop: Unravelling Full Stack Safety in Agentic Collaborative Driving

Xiangbo Gao, Tzu-Hsiang Lin, Ruojing Song et al. (+6 more)

Collaborative driving systems leverage vehicle-to-everything (V2X) communication across multiple agents to enhance driving safety and efficiency. Traditional V2X systems take raw sensor data, neural features, or perception results as communication media, which face persistent challenges, including high bandwidth demands, semantic loss, and interoperability issues. Recent advances investigate natural language as a promising medium, which can provide semantic richness, decision-level reasoning, and human-machine interoperability at significantly lower bandwidth. Despite great promise, this paradigm shift also introduces new vulnerabilities within language communication, including message loss, hallucinations, semantic manipulation, and adversarial attacks. In this work, we present the first systematic study of full-stack safety and security issues in natural-language-based collaborative driving. Specifically, we develop a comprehensive taxonomy of attack strategies, including connection disruption, relay/replay interference, content spoofing, and multi-connection forgery. To mitigate these risks, we introduce an agentic defense pipeline, which we call SafeCoop, that integrates a semantic firewall, language-perception consistency checks, and multi-source consensus, enabled by an agentic transformation function for cross-frame spatial alignment. We systematically evaluate SafeCoop in closed-loop CARLA simulation across 32 critical scenarios, achieving 69.15% driving score improvement under malicious attacks and up to 67.32% F1 score for malicious detection. This study provides guidance for advancing research on safe, secure, and trustworthy language-driven collaboration in transportation systems. Our project page is https://xiangbogaobarry.github.io/SafeCoop.
Read the full abstract

All authors Xiangbo Gao, Tzu-Hsiang Lin, Ruojing Song, Yuheng Wu, Kuan-Ru Huang, Zicheng Jin, Fangzhou Lin, Shinan Liu, Zhengzhong Tu

Computation and Language

Does Reasoning Help LLM Agents Play Dungeons and Dragons? A Prompt Engineering Experiment

Patricia Delafuente, Arya Honraopatil and Lara J. Martin

This paper explores the application of Large Language Models (LLMs) and reasoning to predict Dungeons & Dragons (DnD) player actions and format them as Avrae Discord bot commands. Using the FIREBALL dataset, we evaluated a reasoning model, DeepSeek-R1-Distill-LLaMA-8B, and an instruct model, LLaMA-3.1-8B-Instruct, for command generation. Our findings highlight the importance of providing specific instructions to models, that even single sentence changes in prompts can greatly affect the output of models, and that instruct models are sufficient for this task compared to reasoning models.
Read the full abstract