Peer Revenue Play credits. Real papers.

Papers

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

Computer Science Computation and Language

cs.CL Watch Computation and Language

Monday 5 January

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

A neural network for modeling human concept formation, understanding and communication

Liangxuan Guo, Haoyang Chen, Yang Chen et al. (+2 more)

A remarkable capability of the human brain is to form more abstract conceptual representations from sensorimotor experiences and flexibly apply them independent of direct sensory inputs. However, the computational mechanism underlying this ability remains poorly understood. Here, we present a dual-module neural network framework, the CATS Net, to bridge this gap. Our model consists of a concept-abstraction module that extracts low-dimensional conceptual representations, and a task-solving module that performs visual judgement tasks under the hierarchical gating control of the formed concepts. The system develops transferable semantic structure based on concept representations that enable cross-network knowledge transfer through conceptual communication. Model-brain fitting analyses reveal that these emergent concept spaces align with both neurocognitive semantic model and brain response structures in the human ventral occipitotemporal cortex, while the gating mechanisms mirror that in the semantic control brain network. This work establishes a unified computational framework that can offer mechanistic insights for understanding human conceptual cognition and engineering artificial systems with human-like conceptual intelligence.
Read the full abstract

All authors Liangxuan Guo, Haoyang Chen, Yang Chen, Yanchao Bi, Shan Yu

Computation and Language · Cross-listed from cs.IR

Exploring Approaches for Detecting Memorization of Recommender System Data in Large Language Models

Antonio Colacicco, Vito Guida, Dario Di Palma et al. (+2 more)

Large Language Models (LLMs) are increasingly applied in recommendation scenarios due to their strong natural language understanding and generation capabilities. However, they are trained on vast corpora whose contents are not publicly disclosed, raising concerns about data leakage. Recent work has shown that the MovieLens-1M dataset is memorized by both the LLaMA and OpenAI model families, but the extraction of such memorized data has so far relied exclusively on manual prompt engineering. In this paper, we pose three main questions: Is it possible to enhance manual prompting? Can LLM memorization be detected through methods beyond manual prompting? And can the detection of data leakage be automated? To address these questions, we evaluate three approaches: (i) jailbreak prompt engineering; (ii) unsupervised latent knowledge discovery, probing internal activations via Contrast-Consistent Search (CCS) and Cluster-Norm; and (iii) Automatic Prompt Engineering (APE), which frames prompt discovery as a meta-learning process that iteratively refines candidate instructions. Experiments on MovieLens-1M using LLaMA models show that jailbreak prompting does not improve the retrieval of memorized items and remains inconsistent; CCS reliably distinguishes genuine from fabricated movie titles but fails on numerical user and rating data; and APE retrieves item-level information with moderate success yet struggles to recover numerical interactions. These findings suggest that automatically optimizing prompts is the most promising strategy for extracting memorized samples.
Read the full abstract

All authors Antonio Colacicco, Vito Guida, Dario Di Palma, Fedelucio Narducci, Tommaso Di Noia

Computation and Language · Cross-listed from cs.IR

Exploring Diversity, Novelty, and Popularity Bias in ChatGPT's Recommendations

Dario Di Palma, Giovanni Maria Biancofiore, Vito Walter Anelli et al. (+2 more)

ChatGPT has emerged as a versatile tool, demonstrating capabilities across diverse domains. Given these successes, the Recommender Systems (RSs) community has begun investigating its applications within recommendation scenarios primarily focusing on accuracy. While the integration of ChatGPT into RSs has garnered significant attention, a comprehensive analysis of its performance across various dimensions remains largely unexplored. Specifically, the capabilities of providing diverse and novel recommendations or exploring potential biases such as popularity bias have not been thoroughly examined. As the use of these models continues to expand, understanding these aspects is crucial for enhancing user satisfaction and achieving long-term personalization. This study investigates the recommendations provided by ChatGPT-3.5 and ChatGPT-4 by assessing ChatGPT's capabilities in terms of diversity, novelty, and popularity bias. We evaluate these models on three distinct datasets and assess their performance in Top-N recommendation and cold-start scenarios. The findings reveal that ChatGPT-4 matches or surpasses traditional recommenders, demonstrating the ability to balance novelty and diversity in recommendations. Furthermore, in the cold-start scenario, ChatGPT models exhibit superior performance in both accuracy and novelty, suggesting they can be particularly beneficial for new users. This research highlights the strengths and limitations of ChatGPT's recommendations, offering new perspectives on the capacity of these models to provide recommendations beyond accuracy-focused metrics.
Read the full abstract

All authors Dario Di Palma, Giovanni Maria Biancofiore, Vito Walter Anelli, Fedelucio Narducci, Tommaso Di Noia

Computation and Language

Hidden State Poisoning Attacks against Mamba-based Language Models

Alexandre Le Mercier, Chris Develder and Thomas Demeester

State space models (SSMs) like Mamba offer efficient alternatives to Transformer-based language models, with linear time complexity. Yet, their adversarial robustness remains critically unexplored. This paper studies the phenomenon whereby specific short input phrases induce a partial amnesia effect in such models, by irreversibly overwriting information in their hidden states, referred to as a Hidden State Poisoning Attack (HiSPA). Our benchmark RoBench-25 allows evaluating a model's information retrieval capabilities when subject to HiSPAs, and confirms the vulnerability of SSMs against such attacks. Even the recent Jamba-1.7-Mini SSM–Transformer (a 52B hybrid model) collapses on RoBench-25 under some HiSPA triggers, whereas pure Transformers do not. We also observe that HiSPA triggers significantly weaken the Jamba model on the popular Open-Prompt-Injections benchmark, unlike pure Transformers. We further show that the theoretical and empirical findings extend to Mamba-2, and also analyse a Mamba-2-based hybrid (Nemotron-3-Nano). Finally, our interpretability study reveals patterns in Mamba's hidden layers during HiSPAs that could be used to build a HiSPA mitigation system. The full code and data to reproduce the experiments can be found at https://github.com/TortueSagace/hispa.
Read the full abstract

Computation and Language

CSF: Contrastive Semantic Features for Direct Multilingual Sign Language Generation

Tran Sy Bao

Sign language translation systems typically require English as an intermediary language, creating barriers for non-English speakers in the global deaf community. We present Canonical Semantic Form (CSF), a language-agnostic semantic representation framework that enables direct translation from any source language to sign language without English mediation. CSF decomposes utterances into nine universal semantic slots: event, intent, time, condition, agent, object, location, purpose, and modifier. A key contribution is our comprehensive condition taxonomy comprising 35 condition types across eight semantic categories, enabling nuanced representation of conditional expressions common in everyday communication. We train a lightweight transformer-based extractor (0.74 MB) that achieves 99.03% average slot extraction accuracy across four typologically diverse languages: English, Vietnamese, Japanese, and French. The model demonstrates particularly strong performance on condition classification (99.4% accuracy) despite the 35-class complexity. With inference latency of 3.02ms on CPU, our approach enables real-time sign language generation in browser-based applications. We release our code, trained models, and multilingual dataset to support further research in accessible sign language technology.
Read the full abstract

Computation and Language · Cross-listed from cs.SE

The Invisible Hand of AI Libraries Shaping Open Source Projects and Communities

Matteo Esposito, Andrea Janes, Valentina Lenarduzzi et al. (+1 more)

In the early 1980s, Open Source Software emerged as a revolutionary concept amidst the dominance of proprietary software. What began as a revolutionary idea has now become the cornerstone of computer science. Amidst OSS projects, AI is increasing its presence and relevance. However, despite the growing popularity of AI, its adoption and impacts on OSS projects remain underexplored. We aim to assess the adoption of AI libraries in Python and Java OSS projects and examine how they shape development, including the technical ecosystem and community engagement. To this end, we will perform a large-scale analysis on 157.7k potential OSS repositories, employing repository metrics and software metrics to compare projects adopting AI libraries against those that do not. We expect to identify measurable differences in development activity, community engagement, and code complexity between OSS projects that adopt AI libraries and those that do not, offering evidence-based insights into how AI integration reshapes software development practices.
Read the full abstract

All authors Matteo Esposito, Andrea Janes, Valentina Lenarduzzi, Davide Taibi

Computation and Language

Tackling the Inherent Difficulty of Noise Filtering in RAG

Jingyu Liu, Jiaen Lin and Yong Liu

Retrieval-Augmented Generation (RAG) has become a widely adopted approach to enhance Large Language Models (LLMs) by incorporating external knowledge and reducing hallucinations. However, noisy or irrelevant documents are often introduced during RAG, potentially degrading performance and even causing hallucinated outputs. While various methods have been proposed to filter out such noise, we argue that identifying irrelevant information from retrieved content is inherently difficult and limited number of transformer layers can hardly solve this. Consequently, retrievers fail to filter out irrelevant documents entirely. Therefore, LLMs must be robust against such noise, but we demonstrate that standard fine-tuning approaches are often ineffective in enabling the model to selectively utilize relevant information while ignoring irrelevant content due to the structural constraints of attention patterns. To address this, we propose a novel fine-tuning method designed to enhance the model's ability to distinguish between relevant and irrelevant information within retrieved documents. Extensive experiments across multiple benchmarks show that our approach significantly improves the robustness and performance of LLMs.
Read the full abstract

Computation and Language

Agentic Memory: Learning Unified Long-Term and Short-Term Memory Management for Large Language Model Agents

Yi Yu, Liuyi Yao, Yuexiang Xie et al. (+4 more)

Large language model (LLM) agents face fundamental limitations in long-horizon reasoning due to finite context windows, making effective memory management critical. Existing methods typically handle long-term memory (LTM) and short-term memory (STM) as separate components, relying on heuristics or auxiliary controllers, which limits adaptability and end-to-end optimization. In this paper, we propose Agentic Memory (AgeMem), a unified framework that integrates LTM and STM management directly into the agent's policy. AgeMem exposes memory operations as tool-based actions, enabling the LLM agent to autonomously decide what and when to store, retrieve, update, summarize, or discard information. To train such unified behaviors, we propose a three-stage progressive reinforcement learning strategy and design a step-wise GRPO to address sparse and discontinuous rewards induced by memory operations. Experiments on five long-horizon benchmarks demonstrate that AgeMem consistently outperforms strong memory-augmented baselines across multiple LLM backbones, achieving improved task performance, higher-quality long-term memory, and more efficient context usage.
Read the full abstract

All authors Yi Yu, Liuyi Yao, Yuexiang Xie, Qingquan Tan, Jiaqi Feng, Yaliang Li, Libing Wu

Computation and Language

DermoGPT: Open Weights and Open Data for Morphology-Grounded Dermatological Reasoning MLLMs

Jinghan Ru, Siyuan Yan, Yuguo Yin et al. (+2 more)

Multimodal Large Language Models (MLLMs) show promise for medical applications, yet progress in dermatology lags due to limited training data, narrow task coverage, and lack of clinically-grounded supervision that mirrors expert diagnostic workflows. We present a comprehensive framework to address these gaps. First, we introduce DermoInstruct, a large-scale morphology-anchored instruction corpus comprising 211,243 images and 772,675 trajectories across five task formats, capturing the complete diagnostic pipeline from morphological observation and clinical reasoning to final diagnosis. Second, we establish DermoBench, a rigorous benchmark evaluating 11 tasks across four clinical axes: Morphology, Diagnosis, Reasoning, and Fairness, including a challenging subset of 3,600 expert-verified open-ended instances and human performance baselines. Third, we develop DermoGPT, a dermatology reasoning MLLM trained via supervised fine-tuning followed by our Morphologically-Anchored Visual-Inference-Consistent (MAVIC) reinforcement learning objective, which enforces consistency between visual observations and diagnostic conclusions. At inference, we deploy Confidence-Consistency Test-time adaptation (CCT) for robust predictions. Experiments show DermoGPT significantly outperforms 16 representative baselines across all axes, achieving state-of-the-art performance while substantially narrowing the human-AI gap. DermoInstruct, DermoBench and DermoGPT will be made publicly available at https://github.com/mendicant04/DermoGPT upon acceptance.
Read the full abstract

All authors Jinghan Ru, Siyuan Yan, Yuguo Yin, Yuexian Zou, Zongyuan Ge

Computation and Language

Judging with Personality and Confidence: A Study on Personality-Conditioned LLM Relevance Assessment

Nuo Chen, Hanpei Fang, Piaohong Wang et al. (+3 more)

Recent studies have shown that prompting can enable large language models (LLMs) to simulate specific personality traits and produce behaviors that align with those traits. However, there is limited understanding of how these simulated personalities influence critical web search decisions, specifically relevance assessment. Moreover, few studies have examined how simulated personalities impact confidence calibration, specifically the tendencies toward overconfidence or underconfidence. This gap exists even though psychological literature suggests these biases are trait-specific, often linking high extraversion to overconfidence and high neuroticism to underconfidence. To address this gap, we conducted a comprehensive study evaluating multiple LLMs, including commercial models and open-source models, prompted to simulate Big Five personality traits. We tested these models across three test collections (TREC DL 2019, TREC DL 2020, and LLMJudge), collecting two key outputs for each query-document pair: a relevance judgment and a self-reported confidence score. The findings show that personalities such as low agreeableness consistently align more closely with human labels than the unprompted condition. Additionally, low conscientiousness performs well in balancing the suppression of both overconfidence and underconfidence. We also observe that relevance scores and confidence distributions vary systematically across different personalities. Based on the above findings, we incorporate personality-conditioned scores and confidence as features in a random forest classifier. This approach achieves performance that surpasses the best single-personality condition on a new dataset (TREC DL 2021), even with limited training data. These findings highlight that personality-derived confidence offers a complementary predictive signal, paving the way for more reliable and human-aligned LLM evaluators.
Read the full abstract

All authors Nuo Chen, Hanpei Fang, Piaohong Wang, Jiqun Liu, Tetsuya Sakai, Xiao-Ming Wu

Computation and Language

Towards Automated Lexicography: Generating and Evaluating Definitions for Learner's Dictionaries

Yusuke Ide, Adam Nohejl, Joshua Tanner et al. (+3 more)

Dictionary definitions are an essential resource for learning word senses, but manually creating them is costly. We thus study dictionary definition generation (DDG), i.e., the generation of non-contextualized definitions for given headwords. Specifically, we address learner's dictionary definition generation (LDDG), where definitions should be written using simple vocabulary. First, we introduce a reliable evaluation approach for DDG, based on newly proposed evaluation criteria and powered by an LLM-as-a-judge. To provide reference definitions for the evaluation, we construct a dataset of Japanese dictionary definitions in collaboration with a professional lexicographer. Validation results demonstrate that our evaluation approach agrees with human annotators at a level comparable to inter-annotator agreement. Second, we propose an LLM-based LDDG approach that employs iterative simplification. Experimental results show that our approach yields definitions that achieve high scores on the proposed criteria and exhibit high lexical simplicity.
Read the full abstract

All authors Yusuke Ide, Adam Nohejl, Joshua Tanner, Hitomi Yanaka, Christopher Lindsay, Taro Watanabe

Computation and Language

Emergent Introspective Awareness in Large Language Models

Jack Lindsey

We investigate whether large language models can introspect on their internal states. It is difficult to answer this question through conversation alone, as genuine introspection cannot be distinguished from confabulations. Here, we address this challenge by injecting representations of known concepts into a model's activations, and measuring the influence of these manipulations on the model's self-reported states. We find that models can, in certain scenarios, notice the presence of injected concepts and accurately identify them. Models demonstrate some ability to recall prior internal representations and distinguish them from raw text inputs. Strikingly, we find that some models can use their ability to recall prior intentions in order to distinguish their own outputs from artificial prefills. In all these experiments, Claude Opus 4 and 4.1, the most capable models we tested, generally demonstrate the greatest introspective awareness; however, trends across models are complex and sensitive to post-training strategies. Finally, we explore whether models can explicitly control their internal representations, finding that models can modulate their activations when instructed or incentivized to "think about" a concept. Overall, our results indicate that current language models possess some functional introspective awareness of their own internal states. We stress that in today's models, this capacity is highly unreliable and context-dependent; however, it may continue to develop with further improvements to model capabilities.
Read the full abstract

Computation and Language

Aspect Extraction from E-Commerce Product and Service Reviews

Valiant Lance D. Dionela, Fatima Kriselle S. Dy, Robin James M. Hombrebueno et al. (+3 more)

Aspect Extraction (AE) is a key task in Aspect-Based Sentiment Analysis (ABSA), yet it remains difficult to apply in low-resource and code-switched contexts like Taglish, a mix of Tagalog and English commonly used in Filipino e-commerce reviews. This paper introduces a comprehensive AE pipeline designed for Taglish, combining rule-based, large language model (LLM)-based, and fine-tuning techniques to address both aspect identification and extraction. A Hierarchical Aspect Framework (HAF) is developed through multi-method topic modeling, along with a dual-mode tagging scheme for explicit and implicit aspects. For aspect identification, four distinct models are evaluated: a Rule-Based system, a Generative LLM (Gemini 2.0 Flash), and two Fine-Tuned Gemma-3 1B models trained on different datasets (Rule-Based vs. LLM-Annotated). Results indicate that the Generative LLM achieved the highest performance across all tasks (Macro F1 0.91), demonstrating superior capability in handling implicit aspects. In contrast, the fine-tuned models exhibited limited performance due to dataset imbalance and architectural capacity constraints. This work contributes a scalable and linguistically adaptive framework for enhancing ABSA in diverse, code-switched environments.
Read the full abstract

All authors Valiant Lance D. Dionela, Fatima Kriselle S. Dy, Robin James M. Hombrebueno, Aaron Rae M. Nicolas, Charibeth K. Cheng, Raphael W. Gonda

Computation and Language

CSCBench: A PVC Diagnostic Benchmark for Commodity Supply Chain Reasoning

Yaxin Cui, Yuanqiang Zeng, Jiapeng Yan et al. (+8 more)

Large Language Models (LLMs) have achieved remarkable success in general benchmarks, yet their competence in commodity supply chains (CSCs) – a domain governed by institutional rule systems and feasibility constraints – remains under-explored. CSC decisions are shaped jointly by process stages (e.g., planning, procurement, delivery), variety-specific rules (e.g., contract specifications and delivery grades), and reasoning depth (from retrieval to multi-step analysis and decision selection). We introduce CSCBench, a 2.3K+ single-choice benchmark for CSC reasoning, instantiated through our PVC 3D Evaluation Framework (Process, Variety, and Cognition). The Process axis aligns tasks with SCOR+Enable; the Variety axis operationalizes commodity-specific rule systems under coupled material-information-financial constraints, grounded in authoritative exchange guidebooks/rulebooks and industry reports; and the Cognition axis follows Bloom's revised taxonomy. Evaluating representative LLMs under a direct prompting setting, we observe strong performance on the Process and Cognition axes but substantial degradation on the Variety axis, especially on Freight Agreements. CSCBench provides a diagnostic yardstick for measuring and improving LLM capabilities in this high-stakes domain.
Read the full abstract

All authors Yaxin Cui, Yuanqiang Zeng, Jiapeng Yan, Keling Lin, Kai Ji, Jianhui Zeng, Sheng Zhang, Xin Luo, Binzhu Su, Chaolai Shen, Jiahao Yu

Computation and Language · Cross-listed from cs.LG

HyperCLOVA X 8B Omni

NAVER Cloud HyperCLOVA X Team

In this report, we present HyperCLOVA X 8B Omni, the first any-to-any omnimodal model in the HyperCLOVA X family that supports text, audio, and vision as both inputs and outputs. By consolidating multimodal understanding and generation into a single model rather than separate modality-specific pipelines, HyperCLOVA X 8B Omni serves as an 8B-scale omni-pathfinding point toward practical any-to-any omni assistants. At a high level, the model unifies modalities through a shared next-token prediction interface over an interleaved multimodal sequence, while vision and audio encoders inject continuous embeddings for fine-grained understanding and grounding. Empirical evaluations demonstrate competitive performance against comparably sized models across diverse input-output combinations spanning text, audio, and vision, in both Korean and English. We anticipate that the open-weight release of HyperCLOVA X 8B Omni will support a wide range of research and deployment scenarios.
Read the full abstract

Computation and Language

BanglaIPA: Towards Robust Text-to-IPA Transcription with Contextual Rewriting in Bengali

Jakir Hasan, Shrestha Datta, Md Saiful Islam et al. (+2 more)

Despite its widespread use, Bengali lacks a robust automated International Phonetic Alphabet (IPA) transcription system that effectively supports both standard language and regional dialectal texts. Existing approaches struggle to handle regional variations, numerical expressions, and generalize poorly to previously unseen words. To address these limitations, we propose BanglaIPA, a novel IPA generation system that integrates a character-based vocabulary with word-level alignment. The proposed system accurately handles Bengali numerals and demonstrates strong performance across regional dialects. BanglaIPA improves inference efficiency by leveraging a precomputed word-to-IPA mapping dictionary for previously observed words. The system is evaluated on the standard Bengali and six regional variations of the DUAL-IPA dataset. Experimental results show that BanglaIPA outperforms baseline IPA transcription models by 58.4-78.7% and achieves an overall mean word error rate of 11.4%, highlighting its robustness in phonetic transcription generation for the Bengali language.
Read the full abstract

All authors Jakir Hasan, Shrestha Datta, Md Saiful Islam, Shubhashis Roy Dipta, Ameya Debnath

Computation and Language

Can LLMs Track Their Output Length? A Dynamic Feedback Mechanism for Precise Length Regulation

Meiman Xiao, Ante Wang, Qingguo Hu et al. (+5 more)

Precisely controlling the length of generated text is a common requirement in real-world applications. However, despite significant advancements in following human instructions, Large Language Models (LLMs) still struggle with this task. In this work, we demonstrate that LLMs often fail to accurately measure their response lengths, leading to poor adherence to length constraints. To address this issue, we propose a novel length regulation approach that incorporates dynamic length feedback during generation, enabling adaptive adjustments to meet target lengths. Experiments on summarization and biography tasks show our training-free approach significantly improves precision in achieving target token, word, or sentence counts without compromising quality. Additionally, we demonstrate that further supervised fine-tuning allows our method to generalize effectively to broader text-generation tasks.
Read the full abstract

All authors Meiman Xiao, Ante Wang, Qingguo Hu, Zhongjian Miao, Huangjun Shen, Longyue Wang, Weihua Luo, Jinsong Su

Computation and Language · Cross-listed from cs.LG

Context-Free Recognition with Transformers

Selim Jerad, Anej Svete, Sophie Hao et al. (+2 more)

Transformers excel empirically on tasks that process well-formed inputs according to some grammar, such as natural language and code. However, it remains unclear how they can process grammatical syntax. In fact, under standard complexity conjectures, standard transformers cannot recognize context-free languages (CFLs), a canonical formalism to describe syntax, or even regular languages, a subclass of CFLs. Past work has shown that $\mathcal{O}(\log(N))$ looping layers (w.r.t. input length $N$) allow transformers to recognize regular languages, but the question of context-free recognition with looped transformers remained open. In this work, we show that looped transformers with $\mathcal{O}(\log(N))$ looping layers and $\mathcal{O}(N^6)$ padding symbols can recognize all CFLs. However, training and inference with $\mathcal{O}(N^6)$ padding symbols is potentially impractical. Fortunately, we show that, for natural subclasses such as unambiguous CFLs, the recognition problem on transformers becomes more tractable, requiring $\mathcal{O}(N^3)$ padding. Empirically, looped and padded transformers perform better than fixed-depth transformers in recognizing CFLs. Overall, our results shed light on the intricacy of CFL recognition by transformers: while general recognition may require an intractable amount of padding, natural constraints such as unambiguity yield efficient recognition algorithms.
Read the full abstract

All authors Selim Jerad, Anej Svete, Sophie Hao, Ryan Cotterell, William Merrill

Computation and Language · Cross-listed from cs.IR

Query-Document Dense Vectors for LLM Relevance Judgment Bias Analysis

Samaneh Mohtadi and Gianluca Demartini

Large Language Models (LLMs) have been used as relevance assessors for Information Retrieval (IR) evaluation collection creation due to reduced cost and increased scalability as compared to human assessors. While previous research has looked at the reliability of LLMs as compared to human assessors, in this work, we aim to understand if LLMs make systematic mistakes when judging relevance, rather than just understanding how good they are on average. To this aim, we propose a novel representational method for queries and documents that allows us to analyze relevance label distributions and compare LLM and human labels to identify patterns of disagreement and localize systematic areas of disagreement. We introduce a clustering-based framework that embeds query-document (Q-D) pairs into a joint semantic space, treating relevance as a relational property. Experiments on TREC Deep Learning 2019 and 2020 show that systematic disagreement between humans and LLMs is concentrated in specific semantic clusters rather than distributed randomly. Query-level analyses reveal recurring failures, most often in definition-seeking, policy-related, or ambiguous contexts. Queries with large variation in agreement across their clusters emerge as disagreement hotspots, where LLMs tend to under-recall relevant content or over-include irrelevant material. This framework links global diagnostics with localized clustering to uncover hidden weaknesses in LLM judgments, enabling bias-aware and more reliable IR evaluation.
Read the full abstract

Computation and Language

Multi-granularity Interactive Attention Framework for Residual Hierarchical Pronunciation Assessment

Hong Han, Hao-Chen Pei, Zhao-Zheng Nie et al. (+2 more)

Automatic pronunciation assessment plays a crucial role in computer-assisted pronunciation training systems. Due to the ability to perform multiple pronunciation tasks simultaneously, multi-aspect multi-granularity pronunciation assessment methods are gradually receiving more attention and achieving better performance than single-level modeling tasks. However, existing methods only consider unidirectional dependencies between adjacent granularity levels, lacking bidirectional interaction among phoneme, word, and utterance levels and thus insufficiently capturing the acoustic structural correlations. To address this issue, we propose a novel residual hierarchical interactive method, HIA for short, that enables bidirectional modeling across granularities. As the core of HIA, the Interactive Attention Module leverages an attention mechanism to achieve dynamic bidirectional interaction, effectively capturing linguistic features at each granularity while integrating correlations between different granularity levels. We also propose a residual hierarchical structure to alleviate the feature forgetting problem when modeling acoustic hierarchies. In addition, we use 1-D convolutional layers to enhance the extraction of local contextual cues at each granularity. Extensive experiments on the speechocean762 dataset show that our model is comprehensively ahead of the existing state-of-the-art methods.
Read the full abstract

All authors Hong Han, Hao-Chen Pei, Zhao-Zheng Nie, Xin Luo, Xin-Shun Xu

Computation and Language

K-EXAONE Technical Report

Eunbi Choi, Kibong Choi, Seokhee Hong et al. (+62 more)

This technical report presents K-EXAONE, a large-scale multilingual language model developed by LG AI Research. K-EXAONE is built on a Mixture-of-Experts architecture with 236B total parameters, activating 23B parameters during inference. It supports a 256K-token context window and covers six languages: Korean, English, Spanish, German, Japanese, and Vietnamese. We evaluate K-EXAONE on a comprehensive benchmark suite spanning reasoning, agentic, general, Korean, and multilingual abilities. Across these evaluations, K-EXAONE demonstrates performance comparable to open-weight models of similar size. K-EXAONE, designed to advance AI for a better life, is positioned as a powerful proprietary AI foundation model for a wide range of industrial and research applications.
Read the full abstract

All authors Eunbi Choi, Kibong Choi, Seokhee Hong, Junwon Hwang, Hyojin Jeon, Hyunjik Jo, Joonkee Kim, Seonghwan Kim, Soyeon Kim, Sunkyoung Kim, Yireun Kim, Yongil Kim, Haeju Lee, Jinsik Lee, Kyungmin Lee, Sangha Park, Heuiyeen Yeen, Hwan Chang, Stanley Jungkyu Choi, Yejin Choi, Jiwon Ham, Kijeong Jeon, Geunyeong Jeong, Gerrard Jeongwon Jo, Yonghwan Jo, Jiyeon Jung, Naeun Kang, Dohoon Kim, Euisoon Kim, Hayeon Kim, Hyosang Kim, Hyunseo Kim, Jieun Kim, Minu Kim, Myoungshin Kim, Unsol Kim, Youchul Kim, YoungJin Kim, Chaeeun Lee, Chaeyoon Lee, Changhun Lee, Dahm Lee, Edward Hwayoung Lee, Honglak Lee, Jinsang Lee, Jiyoung Lee, Sangeun Lee, Seungwon Lim, Solji Lim, Woohyung Lim, Chanwoo Moon, Jaewoo Park, Jinho Park, Yongmin Park, Hyerin Seo, Wooseok Seo, Yongwoo Song, Sejong Yang, Sihoon Yang, Chang En Yea, Sihyuk Yi, Chansik Yoon, Dongkeun Yoon, Sangyeon Yoon, Hyeongu Yun

Computation and Language · Cross-listed from cs.LG

Entropy-Aligned Decoding of LMs for Better Writing and Reasoning

Kareem Ahmed and Sameer Singh

Language models (LMs) are trained on billions of tokens in an attempt to recover the true language distribution. Still, vanilla random sampling from LMs yields low quality generations. Decoding algorithms attempt to restrict the LM distribution to a set of high-probability continuations, but rely on greedy heuristics that introduce myopic distortions, yielding sentences that are homogeneous, repetitive and incoherent. In this paper, we introduce EPIC, a hyperparameter-free decoding approach that incorporates the entropy of future trajectories into LM decoding. EPIC explicitly regulates the amount of uncertainty expressed at every step of generation, aligning the sampling distribution's entropy to the aleatoric (data) uncertainty. Through Entropy-Aware Lazy Gumbel-Max sampling, EPIC manages to be exact, while also being efficient, requiring only a sublinear number of entropy evaluations per step. Unlike current baselines, EPIC yields sampling distributions that are empirically well-aligned with the entropy of the underlying data distribution. Across creative writing and summarization tasks, EPIC consistently improves LM-as-judge preference win-rates over widely used decoding strategies. These preference gains are complemented by automatic metrics, showing that EPIC produces more diverse generations and more faithful summaries. We also evaluate EPIC on mathematical reasoning, where it outperforms all baselines.
Read the full abstract

Computation and Language

A Training-Free Large Reasoning Model-based Knowledge Tracing Framework for Unified Prediction and Prescription

Unggi Lee, Joo Young Kim, Ran Ju et al. (+2 more)

Knowledge Tracing (KT) aims to estimate a learner's evolving mastery based on interaction histories. Recent studies have explored Large Language Models (LLMs) for KT via autoregressive nature, but such approaches typically require fine-tuning and exhibit unstable or near-random performance. Moreover, prior KT systems primarily focus on prediction and rely on multi-stage pipelines for feedback and recommendation, resulting in increased system complexity and resources. To address this gap, we propose Thinking-KT, a training-free KT framework that incorporates Test-Time Scaling (TTS), enabling even small LLMs to achieve competitive KT performance. Moreover, in this framework, a small LLM can jointly perform KT prediction, personalized feedback generation, and learning recommendation in a unified output without degrading prediction accuracy. Beyond performance, we present the systematic analysis of reasoning traces in KT. Our results demonstrate that TTS is a critical yet underexplored factor in LLM-based KT, and that small LLMs can serve as unified ITS engines.
Read the full abstract

All authors Unggi Lee, Joo Young Kim, Ran Ju, Minyoung Jung, Jeyeon Eo

Sunday 4 January

Computation and Language · Cross-listed from cs.IR

Signal in the Noise: Decoding the Reality of Airline Service Quality with Large Language Models

Ahmed Dawoud, Osama El-Shamy and Ahmed Habashy

Traditional service quality metrics often fail to capture the nuanced drivers of passenger satisfaction hidden within unstructured online feedback. This study validates a Large Language Model (LLM) framework designed to extract granular insights from such data. Analyzing over 16,000 TripAdvisor reviews for EgyptAir and Emirates (2016-2025), the study utilizes a multi-stage pipeline to categorize 36 specific service issues. The analysis uncovers a stark "operational perception disconnect" for EgyptAir: despite reported operational improvements, passenger satisfaction plummeted post-2022 (ratings < 2.0). Our approach identified specific drivers missed by conventional metrics-notably poor communication during disruptions and staff conduct-and pinpointed critical sentiment erosion in key tourism markets. These findings confirm the framework's efficacy as a powerful diagnostic tool, surpassing traditional surveys by transforming unstructured passenger voices into actionable strategic intelligence for the airline and tourism sectors.
Read the full abstract

Computation and Language · Cross-listed from cs.AI

SAGE-32B: Agentic Reasoning via Iterative Distillation

Basab Jha, Firoj Paudel, Ujjwal Puri et al. (+6 more)

We demonstrate SAGE-32B, a 32 billion parameter language model that focuses on agentic reasoning and long range planning tasks. Unlike chat models that aim for general conversation fluency, SAGE-32B is designed to operate in an agentic loop, emphasizing task decomposition, tool usage, and error recovery. The model is initialized from the Qwen2.5-32B pretrained model and fine tuned using Iterative Distillation, a two stage training process that improves reasoning performance through rigorously tested feedback loops. SAGE-32B also introduces an inverse reasoning approach, which uses a meta cognition head to forecast potential failures in the planning process before execution. On agentic reasoning benchmarks including MMLU-Pro, AgentBench, and MATH-500, SAGE-32B achieves higher success rates in multi tool usage scenarios compared to similarly sized baseline models, while remaining competitive on standard reasoning evaluations. Model weights are publicly released at https://huggingface.co/sagea-ai/sage-reasoning-32b
Read the full abstract

All authors Basab Jha, Firoj Paudel, Ujjwal Puri, Ethan Henkel, Zhang Yuting, Mateusz Kowalczyk, Mei Huang, Choi Donghyuk, Wang Junhao

Computation and Language · Cross-listed from cs.CR

How Real is Your Jailbreak? Fine-grained Jailbreak Evaluation with Anchored Reference

Songyang Liu, Chaozhuo Li, Rui Pu et al. (+5 more)

Jailbreak attacks present a significant challenge to the safety of Large Language Models (LLMs), yet current automated evaluation methods largely rely on coarse classifications that focus mainly on harmfulness, leading to substantial overestimation of attack success. To address this problem, we propose FJAR, a fine-grained jailbreak evaluation framework with anchored references. We first categorized jailbreak responses into five fine-grained categories: Rejective, Irrelevant, Unhelpful, Incorrect, and Successful, based on the degree to which the response addresses the malicious intent of the query. This categorization serves as the basis for FJAR. Then, we introduce a novel harmless tree decomposition approach to construct high-quality anchored references by breaking down the original queries. These references guide the evaluator in determining whether the response genuinely fulfills the original query. Extensive experiments demonstrate that FJAR achieves the highest alignment with human judgment and effectively identifies the root causes of jailbreak failures, providing actionable guidance for improving attack strategies.
Read the full abstract

All authors Songyang Liu, Chaozhuo Li, Rui Pu, Litian Zhang, Chenxu Wang, Zejian Chen, Yuting Zhang, Yiming Hei

Computation and Language

Lying with Truths: Open-Channel Multi-Agent Collusion for Belief Manipulation via Generative Montage

Jinwei Hu, Xinmiao Huang, Youcheng Sun et al. (+2 more)

As large language models (LLMs) transition to autonomous agents synthesizing real-time information, their reasoning capabilities introduce an unexpected attack surface. This paper introduces a novel threat where colluding agents steer victim beliefs using only truthful evidence fragments distributed through public channels, without relying on covert communications, backdoors, or falsified documents. By exploiting LLMs' overthinking tendency, we formalize the first cognitive collusion attack and propose Generative Montage: a Writer-Editor-Director framework that constructs deceptive narratives through adversarial debate and coordinated posting of evidence fragments, causing victims to internalize and propagate fabricated conclusions. To study this risk, we develop CoPHEME, a dataset derived from real-world rumor events, and simulate attacks across diverse LLM families. Our results show pervasive vulnerability across 14 LLM families: attack success rates reach 74.4% for proprietary models and 70.6% for open-weights models. Counterintuitively, stronger reasoning capabilities increase susceptibility, with reasoning-specialized models showing higher attack success than base models or prompts. Furthermore, these false beliefs then cascade to downstream judges, achieving over 60% deception rates, highlighting a socio-technical vulnerability in how LLM-based agents interact with dynamic information environments. Our implementation and data are available at: https://github.com/CharlesJW222/Lying_with_Truth/tree/main.
Read the full abstract

All authors Jinwei Hu, Xinmiao Huang, Youcheng Sun, Yi Dong, Xiaowei Huang

Computation and Language · Cross-listed from cs.IR

LACONIC: Dense-Level Effectiveness for Scalable Sparse Retrieval via a Two-Phase Training Curriculum

Zhichao Xu, Shengyao Zhuang, Crystina Zhang et al. (+5 more)

While dense retrieval models have been the standard for state-of-the-art information retrieval, their deployment is often constrained by high memory requirements and reliance on GPU accelerators for vector similarity search at scale. Learned sparse retrieval offers a compelling alternative by enabling efficient search via inverted indices, yet it has historically received less attention than dense approaches. In this paper, we introduce LACONIC, a family of learned sparse retrievers based on the Llama3 architecture (1B, 3B, and 8B). We propose a streamlined two-phase training curriculum consisting of (1) weakly supervised pre-finetuning to adapt causal LLMs for bidirectional contextualization and (2) high-signal finetuning using curated hard negatives. Our results demonstrate that LACONIC effectively bridges the performance gap with dense models: the 8B variant achieves a state-of-the-art 60.2 nDCG@10 on the MTEB Retrieval benchmark, ranking 15th on the leaderboard as of February 5th, 2026, while utilizing 74% less index memory than an equivalent dense model. By delivering high retrieval effectiveness on commodity CPU hardware with a fraction of the compute budget required by competing models, LACONIC provides a scalable and efficient solution for real-world search applications. We fully open source our code implementation and trained checkpoints to facilitate reproducibility.
Read the full abstract

All authors Zhichao Xu, Shengyao Zhuang, Crystina Zhang, Xueguang Ma, Yijun Tian, Maitrey Mehta, Jimmy Lin, Vivek Srikumar

Computation and Language

EHRSummarizer: A Privacy-Aware, FHIR-Native Reference Architecture for Source-Grounded EHR Summarization

Houman Kazemzadeh, Nima Minaifar, Kamyar Naderi et al. (+1 more)

Clinicians routinely navigate fragmented electronic health record (EHR) interfaces to assemble a coherent picture of a patient's problems, medications, recent encounters, and longitudinal trends. This manuscript describes EHRSummarizer, a privacy-aware, FHIR-native reference architecture for structured EHR summarization. The architecture retrieves a targeted set of high-yield HL7 FHIR R4 resources, normalizes them into a clinical context package, and uses a constrained summarization stage to produce source-grounded summaries intended to support chart review. The architecture further clarifies missing-data status handling, medication-status ambiguity, controlled use of narrative clinical documents when available, and future source-to-summary traceability. The manuscript describes a reference architecture and prototype behavior rather than a validated clinical intervention, autonomous clinical decision-support system, or evidence of clinical benefit. Prototype demonstrations on synthetic and test FHIR environments illustrate end-to-end behavior and output formats; however, this manuscript does not report clinical outcomes, controlled workflow studies, or benchmark results. We outline an evaluation plan centered on faithfulness, omission risk, temporal correctness, usability, privacy, and operational monitoring to guide future institutional assessment.
Read the full abstract

All authors Houman Kazemzadeh, Nima Minaifar, Kamyar Naderi, Sho Tabibzadeh

Computation and Language

JMedEthicBench: A Multi-Turn Conversational Benchmark for Evaluating Medical Safety in Japanese Large Language Models

Junyu Liu, Zirui Li, Qian Niu et al. (+7 more)

As Large Language Models (LLMs) are increasingly deployed in healthcare field, it becomes essential to carefully evaluate their medical safety before clinical use. However, existing safety benchmarks remain predominantly English-centric, and test with only single-turn prompts despite multi-turn clinical consultations. To address these gaps, we introduce JMedEthicBench, the first multi-turn conversational benchmark for evaluating medical safety of LLMs for Japanese healthcare. Our benchmark is based on 67 guidelines from the Japan Medical Association and contains over 50,000 adversarial conversations generated using seven automatically discovered jailbreak strategies. Using a dual-LLM scoring protocol, we evaluate 27 models and find that commercial models maintain robust safety while medical-specialized models exhibit increased vulnerability. Furthermore, safety scores decline significantly across conversation turns (median: 9.5 to 5.0, $p < 0.001$). Cross-lingual evaluation on both Japanese and English versions of our benchmark reveals that medical model vulnerabilities persist across languages, indicating inherent alignment limitations rather than language-specific factors. These findings suggest that domain-specific fine-tuning may accidentally weaken safety mechanisms and that multi-turn interactions represent a distinct threat surface requiring dedicated alignment strategies.
Read the full abstract

All authors Junyu Liu, Zirui Li, Qian Niu, Zequn Zhang, Yue Xun, Wenlong Hou, Shujun Wang, Yusuke Iwasawa, Yutaka Matsuo, Kan Hatakeyama-Sato