Peer Revenue Publish or profit

Papers

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

Computer Science Computation and Language

cs.CL Watch Computation and Language

Thursday 23 April

Computation and Language

TingIS: Real-time Risk Event Discovery from Noisy Customer Incidents at Enterprise Scale

Jun Wang, Ziyin Zhang, Rui Wang et al. (+3 more)

Real-time detection and mitigation of technical anomalies are critical for large-scale cloud-native services, where even minutes of downtime can result in massive financial losses and diminished user trust. While customer incidents serve as a vital signal for discovering risks missed by monitoring, extracting actionable intelligence from this data remains challenging due to extreme noise, high throughput, and semantic complexity of diverse business lines. In this paper, we present TingIS, an end-to-end system designed for enterprise-grade incident discovery. At the core of TingIS is a multi-stage event linking engine that synergizes efficient indexing techniques with Large Language Models (LLMs) to make informed decisions on event merging, enabling the stable extraction of actionable incidents from just a handful of diverse user descriptions. This engine is complemented by a cascaded routing mechanism for precise business attribution and a multi-dimensional noise reduction pipeline that integrates domain knowledge, statistical patterns, and behavioral filtering. Deployed in a production environment handling a peak throughput of over 2,000 messages per minute and 300,000 messages per day, TingIS achieves a P90 alert latency of 3.5 minutes and a 95% discovery rate for high-priority incidents. Benchmarks constructed from real-world data demonstrate that TingIS significantly outperforms baseline methods in routing accuracy, clustering quality, and Signal-to-Noise Ratio.
Read the full abstract

All authors Jun Wang, Ziyin Zhang, Rui Wang, Hang Yu, Peng Di, Rui Wang

Computation and Language

A Multimodal Text- and Graph-Based Approach for Open-Domain Event Extraction from Documents

Praval Sharma

Event extraction is essential for event understanding and analysis. It supports tasks such as document summarization and decision-making in emergency scenarios. However, existing event extraction approaches have limitations: (1) closed-domain algorithms are restricted to predefined event types and thus rarely generalize to unseen types and (2) open-domain event extraction algorithms, capable of handling unconstrained event types, have largely overlooked the potential of large language models (LLMs) despite their advanced abilities. Additionally, they do not explicitly model document-level contextual, structural, and semantic reasoning, which are crucial for effective event extraction but remain challenging for LLMs due to lost-in-the-middle phenomenon and attention dilution. To address these limitations, we propose multimodal open-domain event extraction, MODEE , a novel approach for open-domain event extraction that combines graph-based learning with text-based representation from LLMs to model document-level reasoning. Empirical evaluations on large datasets demonstrate that MODEE outperforms state-of-the-art open-domain event extraction approaches and can be generalized to closed-domain event extraction, where it outperforms existing algorithms.
Read the full abstract

Computation and Language

Revisiting Non-Verbatim Memorization in Large Language Models: The Role of Entity Surface Forms

Yuto Nishida, Naoki Shikoda, Yosuke Kishinami et al. (+4 more)

Understanding what kinds of factual knowledge large language models (LLMs) memorize is essential for evaluating their reliability and limitations. Entity-based QA is a common framework for analyzing non-verbatim memorization, but typical evaluations query each entity using a single canonical surface form, making it difficult to disentangle fact memorization from access through a particular name. We introduce RedirectQA, an entity-based QA dataset that uses Wikipedia redirect information to associate Wikidata factual triples with categorized surface forms for each entity, including alternative names, abbreviations, spelling variants, and common erroneous forms. Across 13 LLMs, we examine surface-conditioned factual memorization and find that prediction outcomes often change when only the entity surface form changes. This inconsistency is category-dependent: models are more robust to minor orthographic variations than to larger lexical variations such as aliases and abbreviations. Frequency analyses further suggest that both entity- and surface-level frequencies are associated with accuracy, and that entity frequency often contributes beyond surface frequency. Overall, factual memorization appears neither purely surface-specific nor fully surface-invariant, highlighting the importance of surface-form diversity in evaluating non-verbatim memorization.
Read the full abstract

All authors Yuto Nishida, Naoki Shikoda, Yosuke Kishinami, Ryo Fujii, Makoto Morishita, Hidetaka Kamigaito, Taro Watanabe

Computation and Language

Machine Behavior in Relational Moral Dilemmas: Moral Rightness, Predicted Human Behavior, and Model Decisions

Jiseon Kim, Jea Kwon, Luiz Felipe Vecchietti et al. (+3 more)

Human moral judgment is context-dependent and modulated by interpersonal relationships. As large language models (LLMs) increasingly function as decision-support systems, determining whether they encode these social nuances is critical. We characterize machine behavior using the Whistleblower's Dilemma by varying two experimental dimensions: crime severity and relational closeness. Our study evaluates three distinct perspectives: (1) moral rightness (prescriptive norms), (2) predicted human behavior (descriptive social expectations), and (3) autonomous model decision-making. By analyzing the reasoning processes, we identify a clear cross-perspective divergence: while moral rightness remains consistently fairness-oriented, predicted human behavior shifts significantly toward loyalty as relational closeness increases. Crucially, model decisions align with moral rightness judgments rather than their own behavioral predictions. This inconsistency suggests that LLM decision-making prioritizes rigid, prescriptive rules over the social sensitivity present in their internal world-modeling, which poses a gap that may lead to significant misalignments in real-world deployments.
Read the full abstract

All authors Jiseon Kim, Jea Kwon, Luiz Felipe Vecchietti, Wenchao Dong, Jaehong Kim, Meeyoung Cha

Computation and Language · Cross-listed from cs.AI

Learning to Communicate: Toward End-to-End Optimization of Multi-Agent Language Systems

Ye Yu, Heming Liu, Haibo Jin et al. (+3 more)

Multi-agent systems built on large language models have shown strong performance on complex reasoning tasks, yet most work focuses on agent roles and orchestration while treating inter-agent communication as a fixed interface. Latent communication through internal representations such as key-value caches offers a promising alternative to text-based protocols, but existing approaches do not jointly optimize communication with multi-agent reasoning. Therefore we propose DiffMAS, a training framework that treats latent communication as a learnable component of multi-agent systems. DiffMAS performs parameter-efficient supervised training over multi-agent latent trajectories, enabling agents to jointly learn how information should be encoded and interpreted across interactions. Experiments on mathematical reasoning, scientific QA, code generation, and commonsense benchmarks show that DiffMAS consistently improves reasoning accuracy and decoding stability over single-agent inference, text-based multi-agent systems, and prior latent communication methods, achieving 26.7% on AIME24, 20.2% on GPQA-Diamond, and consistent gains across reasoning benchmarks.
Read the full abstract

All authors Ye Yu, Heming Liu, Haibo Jin, Xiaopeng Yuan, Peng Kuang, Haohan Wang

Computation and Language

SemEval-2026 Task 4: Narrative Story Similarity and Narrative Representation Learning

Hans Ole Hatzel, Ekaterina Artemova, Haimo Paul Stiemer et al. (+2 more)

We present the shared task on narrative similarity and narrative representation learning - NSNRL (pronounced "nass-na-rel"). The task operationalizes narrative similarity as a binary classification problem: determining which of two stories is more similar to an anchor story. We introduce a novel definition of narrative similarity, compatible with both narrative theory and intuitive judgment. Based on the similarity judgments collected under this concept, we also evaluate narrative embedding representations. We collected at least two annotations each for more than 1,000 story summary triples, with each annotation being backed by at least two annotators in agreement. This paper describes the sampling and annotation process for the dataset; further, we give an overview of the submitted systems and the techniques they employ. We received a total of 71 final submissions from 46 teams across our two tracks. In our triple-based classification setup, LLM ensembles make up many of the top-scoring systems, while in the embedding setup, systems with pre- and post-processing on pretrained embedding models perform about on par with custom fine-tuned solutions. Our analysis identifies potential headroom for improvement of automated systems in both tracks. The task website includes visualizations of embeddings alongside instance-level classification results for all teams.
Read the full abstract

All authors Hans Ole Hatzel, Ekaterina Artemova, Haimo Paul Stiemer, Evelyn Gius, Chris Biemann

Computation and Language

Misinformation Span Detection in Videos via Audio Transcripts

Breno Matos, Rennan C. Lima, Savvas Zannettou et al. (+2 more)

Online misinformation is one of the most challenging issues lately, yielding severe consequences, including political polarization, attacks on democracy, and public health risks. Misinformation manifests in any platform with a large user base, including online social networks and messaging apps. It permeates all media and content forms, including images, text, audio, and video. Distinctly, video-based misinformation represents a multifaceted challenge for fact-checkers, given the ease with which individuals can record and upload videos on various video-sharing platforms. Previous research efforts investigated detecting video-based misinformation, focusing on whether a video shares misinformation or not on a video level. While this approach is useful, it only provides a limited and non-easily interpretable view of the problem given that it does not provide an additional context of when misinformation occurs within videos and what content (i.e., claims) are responsible for the video's misinformation nature. In this work, we attempt to bridge this research gap by creating two novel datasets that allow us to explore misinformation detection on videos via audio transcripts, focusing on identifying the span of videos that are responsible for the video's misinformation claim (misinformation span detection). We present two new datasets for this task. We transcribe each video's audio to text, identifying the video segment in which the misinformation claims appears, resulting in two datasets of more than 500 videos with over 2,400 segments containing annotated fact-checked claims. Then, we employ classifiers built with state-of-the-art language models, and our results show that we can identify in which part of a video there is misinformation with an F1 score of 0.68. We make publicly available our annotated datasets. We also release all transcripts, audio and videos.
Read the full abstract

All authors Breno Matos, Rennan C. Lima, Savvas Zannettou, Fabricio Benevenuto, Rodrygo L.T. Santos

Computation and Language

AUDITA: A New Dataset to Audit Humans vs. AI Skill at Audio QA

Tasnim Kabir, Dmytro Kurdydyk, Aadi Palnitkar et al. (+3 more)

Existing audio question answering benchmarks largely emphasize sound event classification or caption-grounded queries, often enabling models to succeed through shortcut strategies, short-duration cues, lexical priors, dataset-specific biases, or even bypassing audio via metadata and captions rather than genuine reasoning Thus, we present AUDITA (Audio Understanding from Diverse Internet Trivia Authors), a large-scale, real-world benchmark to rigorously evaluate audio reasoning beyond surface-level acoustic recognition. AUDITA comprises carefully curated, human-authored trivia questions grounded in real-world audio, designed to stress robust auditory reasoning through challenging distractors and long-range temporal dependencies, using probing queries that cannot be answered from isolated text or sound cues alone. Human average accuracy of 32.13% shows both the challenge of the task while demonstrating meaningful comprehension of the audio. In stark contrast, state of-the-art audio question answering models perform poorly, with average accuracy below 8.86%. Beyond raw accuracy, we apply Item Response Theory (IRT) to estimate latent proficiency, question difficulty, and expose systematic deficiencies of the models and data.
Read the full abstract

All authors Tasnim Kabir, Dmytro Kurdydyk, Aadi Palnitkar, Liam Dorn, Ahmed Haj Ahmed, Jordan Lee Boyd-Graber

Computation and Language

Why are all LLMs Obsessed with Japanese Culture? On the Hidden Cultural and Regional Biases of LLMs

Joseba Fernandez de Landa, Carla Perez-Almendros and Jose Camacho-Collados

LLMs have limitations when it comes to cultural coverage and competence, and in some cases, show specific cultural biases. Although prior studies have examined the cultural capabilities of LLMs, none have specifically investigated their regional preferences in generic culture-related questions. In this work, we propose a new dataset based on a comprehensive taxonomy of Culture-Related Open Questions (CROQ), with questions available in 24 languages. We evaluate LLMs by prompting them to answer questions from CROQ and provide a sample location. The results show that, contrary to previous cultural bias work, LLMs show a clear tendency towards countries such as Japan in their answers. Moreover, our results show that when prompting in languages such as English or other high-resource ones, LLMs tend to provide more diverse outputs. Low-resource languages, on the other hand, show more inclinations towards answering questions highlighting countries for which the input language is an official language. Finally, we also investigate at which point of LLM training this cultural bias emerges, with our results suggesting that the first clear signs appear after supervised fine-tuning, and not during pre-training. Dataset available at https://huggingface.co/datasets/HiTZ/CROQ
Read the full abstract

Computation and Language

StructMem: Structured Memory for Long-Horizon Behavior in LLMs

Buqiang Xu, Yijun Chen, Jizhan Fang et al. (+5 more)

Long-term conversational agents need memory systems that capture relationships between events, not merely isolated facts, to support temporal reasoning and multi-hop question answering. Current approaches face a fundamental trade-off: flat memory is efficient but fails to model relational structure, while graph-based memory enables structured reasoning at the cost of expensive and fragile construction. To address these issues, we propose StructMem, a structure-enriched hierarchical memory framework that preserves event-level bindings and induces cross-event connections. By temporally anchoring dual perspectives and performing periodic semantic consolidation, StructMem improves temporal reasoning and multi-hop performance on LoCoMo, while substantially reducing token usage, API calls, and runtime compared to prior memory systems, see https://github.com/zjunlp/LightMem .
Read the full abstract

All authors Buqiang Xu, Yijun Chen, Jizhan Fang, Ruobin Zhong, Yunzhi Yao, Yuqi Zhu, Lun Du, Shumin Deng

Computation and Language

AEL: Evolving Agent Harness in Open-Ended Environments

Wujiang Xu, Jiaojiao Han, Minghao Guo et al. (+4 more)

LLM Agents Harnesses are hand-designed and stay fixed, so agents accumulate experience but never learn how to use it: which memories to retrieve, when retrieved evidence is misleading, and when the retrieval strategy itself should change. We introduce Agent Evolving Learning (AEL), a two-timescale framework that evolves the harness, recasting memory use as online policy selection. A fast Thompson Sampling bandit selects among memory-retrieval policies episode by episode, while slow LLM reflection follows a diagnose-before-prescribe principle: it first explains why performance degraded, then injects a targeted new retrieval policy as a bandit arm when the current pool plateaus. AEL outperforms ten self-improving and nonLLM baselines on a sequential portfolio benchmark, lifting Sharpe by 27% over the strongest memory-only variant with the lowest variance among all stochastic methods, and generalizes to a support-ticket routing stream, where it improves accuracy by 18% over reflectionfree Thompson Sampling and by 51% over the best prior baseline. Mechanism studies further show that the gains are causal: reflection helps precisely when regimes demand different retrieval behavior, and is provably no-harm/nogain when the best policy is stable. Code and data: https://github.com/WujiangXu/AEL.
Read the full abstract

All authors Wujiang Xu, Jiaojiao Han, Minghao Guo, Kai Mei, Xi Zhu, Han Zhang, Dimitris N. Metaxas

Computation and Language

Beyond N-gram: Data-Aware X-GRAM Extraction for Efficient Embedding Parameter Scaling

Yilong Chen, Yanxi Xie, Zitian Gao et al. (+10 more)

Large token-indexed lookup tables provide a compute-decoupled scaling path, but their practical gains are often limited by poor parameter efficiency and rapid memory growth. We attribute these limitations to Zipfian under-training of the long tail, heterogeneous demand across layers, and "slot collapse" that produces redundant embeddings. To address this, we propose X-GRAM, a frequency-aware dynamic token-injection framework. X-GRAM employs hybrid hashing and alias mixing to compress the tail while preserving head capacity, and refines retrieved vectors via normalized SwiGLU ShortConv to extract diverse local n-gram features. These signals are integrated into attention value streams and inter-layer residuals using depth-aware gating, effectively aligning static memory with dynamic context. This design introduces a memory-centric scaling axis that decouples model capacity from FLOPs. Extensive evaluations at the 0.73B and 1.15B scales show that X-GRAM improves average accuracy by as much as 4.4 points over the vanilla backbone and 3.2 points over strong retrieval baselines, while using substantially smaller tables in the 50% configuration. Overall, by decoupling capacity from compute through efficient memory management, X-GRAM offers a scalable and practical paradigm for future memory-augmented architectures. Code aviliable in https://github.com/Longyichen/X-gram.
Read the full abstract

All authors Yilong Chen, Yanxi Xie, Zitian Gao, He Xin, Yihao Xiao, Jason Klein Liu, Haoming Luo, Yifan Luo, Zhengmao Ye, Tingwen Liu, Xin Zhao, Ran Tao, Bryan Dai

Computation and Language

From If-Statements to ML Pipelines: Revisiting Bias in Code-Generation

Minh Duc Bui, Xenia Heilmann, Mattia Cerrato et al. (+2 more)

Prior work evaluates code generation bias primarily through simple conditional statements, which represent only a narrow slice of real-world programming and reveal solely overt, explicitly encoded bias. We demonstrate that this approach dramatically underestimates bias in practice by examining a more realistic task: generating machine learning (ML) pipelines. Testing both code-specialized and general-instruction large language models, we find that generated pipelines exhibit significant bias during feature selection. Sensitive attributes appear in 87.7% of cases on average, despite models demonstrably excluding irrelevant features (e.g., including "race" while dropping "favorite color" for credit scoring). This bias is substantially more prevalent than that captured by conditional statements, where sensitive attributes appear in only 59.2% of cases. These findings are robust across prompt mitigation strategies, varying numbers of attributes, and different pipeline difficulty levels. Our results challenge simple conditionals as valid proxies for bias evaluation and suggest current benchmarks underestimate bias risk in practical deployments.
Read the full abstract

All authors Minh Duc Bui, Xenia Heilmann, Mattia Cerrato, Manuel Mager, Katharina von der Wense

Computation and Language

Phonological Subspace Collapse Is Aetiology-Specific and Cross-Lingually Stable: Evidence from 3,374 Speakers

Bernard Muller, Antonio Armando Ortiz Barrañón and LaVonne Roberts

We previously introduced a training-free method for dysarthria severity assessment based on d-prime separability of phonological feature subspaces in frozen self-supervised speech representations, validated on 890 speakers across 5 languages with HuBERT-base. Here, we scale the analysis to 3,374 speakers from 25 datasets spanning 12 languages and 5 aetiologies (Parkinson's disease, cerebral palsy, ALS, Down syndrome, and stroke), plus healthy controls, using 6 SSL backbones. We report three findings. First, aetiology-specific degradation profiles are distinguishable at the group level: 10 of 13 features yield large effect sizes (epsilon-squared > 0.14, Holm-corrected p < 0.001), with Parkinson's disease separable from the articulatory execution group at Cohen's d = 0.83; individual-level classification remains limited (22.6% macro F1). Second, profiles show cross-lingual profile-shape stability: cosine similarity of 5-dimensional consonant d-prime profiles exceeds 0.95 across the languages available for each aetiology. Absolute d-prime magnitudes are not cross-lingually calibrated, so the method supports language-independent phenotyping of degradation patterns but requires within-corpus calibration for absolute severity interpretation. Third, the method is architecture-independent: all 6 backbones produce monotonic severity gradients with inter-model agreement exceeding rho = 0.77. Fixed-token d-prime estimation preserves the severity correlation (rho = -0.733 at 200 tokens per class), confirming that the signal is not a token-count artefact. These results support phonological subspace analysis as a robust, training-free framework for aetiology-aware dysarthria characterisation, with evidence of cross-lingual profile-shape stability and cross-backbone robustness in the represented sample.
Read the full abstract

Computation and Language · Cross-listed from cs.CR

Stealthy Backdoor Attacks against LLMs Based on Natural Style Triggers

Jiali Wei, Ming Fan, Guoheng Sun et al. (+3 more)

The growing application of large language models (LLMs) in safety-critical domains has raised urgent concerns about their security. Many recent studies have demonstrated the feasibility of backdoor attacks against LLMs. However, existing methods suffer from three key shortcomings: explicit trigger patterns that compromise naturalness, unreliable injection of attacker-specified payloads in long-form generation, and incompletely specified threat models that obscure how backdoors are delivered and activated in practice. To address these gaps, we present BadStyle, a complete backdoor attack framework and pipeline. BadStyle leverages an LLM as a poisoned sample generator to construct natural and stealthy poisoned samples that carry imperceptible style-level triggers while preserving semantics and fluency. To stabilize payload injection during fine-tuning, we design an auxiliary target loss that reinforces the attacker-specified target content in responses to poisoned inputs and penalizes its emergence in benign responses. We further ground the attack in a realistic threat model and systematically evaluate BadStyle under both prompt-induced and PEFT-based injection strategies. Extensive experiments across seven victim LLMs, including LLaMA, Phi, DeepSeek, and GPT series, demonstrate that BadStyle achieves high attack success rates (ASRs) while maintaining strong stealthiness. The proposed auxiliary target loss substantially improves the stability of backdoor activation, yielding an average ASR improvement of around 30% across style-level triggers. Even in downstream deployment scenarios unknown during injection, the implanted backdoor remains effective. Moreover, BadStyle consistently evades representative input-level defenses and bypasses output-level defenses through simple camouflage.
Read the full abstract

All authors Jiali Wei, Ming Fan, Guoheng Sun, Xicheng Zhang, Haijun Wang, Ting Liu

Computation and Language

Fixation Sequences as Time Series: A Topological Approach to Dyslexia Detection

Marius Huber, David R. Reich and Lena A. Jäger

Persistent homology, a method from topological data analysis, extracts robust, multi-scale features from data. It produces stable representations of time series by applying varying thresholds to their values (a process known as a filtration). We develop novel filtrations for time series and introduce topological methods for the analysis of eye-tracking data, by interpreting fixation sequences as time series, and constructing “hybrid models” that combine topological features with traditional statistical features. We empirically evaluate our method by applying it to the task of dyslexia detection from eye-tracking-while-reading data using the Copenhagen Corpus, which contains scanpaths from dyslexic and non-dyslexic L1 and L2 readers. Our hybrid models outperform existing approaches that rely solely on traditional features, showing that persistent homology captures complementary information encoded in fixation sequences. The strength of these topological features is further underscored by their achieving performance comparable to established baseline methods. Importantly, our proposed filtrations outperform existing ones.
Read the full abstract

Computation and Language

Fine-Grained Perspectives: Modeling Explanations with Annotator-Specific Rationales

Olufunke O. Sarumi, Charles Welch and Daniel Braun

Beyond exploring disaggregated labels for modeling perspectives, annotator rationales provide fine-grained signals of individual perspectives. In this work, we propose a framework for jointly modeling annotator-specific label prediction and corresponding explanations, fine-tuned on the annotators' provided rationales. Using a dataset with disaggregated natural language inference (NLI) annotations and annotator-provided explanations, we condition predictions on both annotator identity and demographic metadata through a representation-level User Passport mechanism. We further introduce two explainer architectures: a post-hoc prompt-based explainer and a prefixed bridge explainer that transfers annotator-conditioned classifier representations directly into a generative model. This design enables explanation generation aligned with individual annotator perspectives. Our results show that incorporating explanation modeling substantially improves predictive performance over a baseline annotator-aware classifier, with the prefixed bridge approach achieving more stable label alignment and higher semantic consistency, while the post-hoc approach yields stronger lexical similarity. These findings indicate that modeling explanations as expressions of fine-grained perspective provides a richer and more faithful representation of disagreement. The proposed approaches advance perspectivist modeling by integrating annotator-specific rationales into both predictive and generative components.
Read the full abstract

Computation and Language · Cross-listed from cs.AI

GS-Quant: Granular Semantic and Generative Structural Quantization for Knowledge Graph Completion

Qizhuo Xie, Yunhui Liu, Yu Xing et al. (+4 more)

Large Language Models (LLMs) have shown immense potential in Knowledge Graph Completion (KGC), yet bridging the modality gap between continuous graph embeddings and discrete LLM tokens remains a critical challenge. While recent quantization-based approaches attempt to align these modalities, they typically treat quantization as flat numerical compression, resulting in semantically entangled codes that fail to mirror the hierarchical nature of human reasoning. In this paper, we propose GS-Quant, a novel framework that generates semantically coherent and structurally stratified discrete codes for KG entities. Unlike prior methods, GS-Quant is grounded in the insight that entity representations should follow a linguistic coarse-to-fine logic. We introduce a Granular Semantic Enhancement module that injects hierarchical knowledge into the codebook, ensuring that earlier codes capture global semantic categories while later codes refine specific attributes. Furthermore, a Generative Structural Reconstruction module imposes causal dependencies on the code sequence, transforming independent discrete units into structured semantic descriptors. By expanding the LLM vocabulary with these learned codes, we enable the model to reason over graph structures isomorphically to natural language generation. Experimental results demonstrate that GS-Quant significantly outperforms existing text-based and embedding-based baselines. Our code is publicly available at https://github.com/mikumifa/GS-Quant.
Read the full abstract

All authors Qizhuo Xie, Yunhui Liu, Yu Xing, Qianzi Hou, Xudong Jin, Tao Zheng, Tieke He

Computation and Language

Multilinguality at the Edge: Developing Language Models for the Global South

Lester James V. Miranda, Songbo Hu, Roi Reichart et al. (+1 more)

Where and how language models (LMs) are deployed determines who can benefit from them. However, there are several challenges that prevent effective deployment of LMs in non-English-speaking and hardware constrained communities in the Global South. We call this challenge the last mile: the intersection of multilinguality and edge deployment, where the goals are aligned but the technical requirements often compete. Studying these two fields together is both a need, as linguistically diverse communities often face the most severe infrastructure constraints, and an opportunity, as edge and multilingual NLP research remain largely siloed. To understand the state of the art and the challenges of combining the two areas, we survey 232 papers that tackle this problem across the language modelling pipeline, from data collection to development and deployment. We also discuss open questions and provide actionable recommendations for different stakeholders in the NLP ecosystem. Finally, we hope that this work contributes to the development of inclusive and equitable language technologies.
Read the full abstract

All authors Lester James V. Miranda, Songbo Hu, Roi Reichart, Anna Korhonen

Computation and Language

Process Supervision via Verbal Critique Improves Reasoning in Large Language Models

Hao-Yuan Chen

Inference-time scaling for LLM reasoning has focused on three axes: chain depth, sample breadth, and learned step-scorers (PRMs). We introduce a fourth axis, granularity of external verbal supervision, via Verbal Process Supervision (VPS), a training-free framework that uses structured natural-language critique from a stronger supervisor to guide an iterative generate-critique-refine loop up to a round budget R. Across GPQA Diamond, AIME 2025, and LiveCodeBench V6 (covering both closed and open models), VPS yields three key results. First, on GPQA Diamond, GPT-5.4 (High) | GPT-5.4 (Low) reaches 94.9% at R=4, surpassing the 94.1% state of the art without gradient updates. Second, on AIME 2025, VPS enables strong weak-actor rescue, boosting scores from 11.7-26.7% to 63.3-90.0% (up to +63.3 points). Third, at matched compute, VPS outperforms Reflexion by +8.5 to +12.1 points and Self-Consistency@5 by +5.0 pp (GPQA) and +8.3 pp (LiveCodeBench), isolating critique granularity as the key driver. Performance scales with the supervisor-actor capability gap (Pearson r=0.90) and degrades when errors are not linguistically expressible (e.g., code synthesis), motivating hybrid verbal-executable methods. These results establish critique granularity as a new axis of inference-time scaling.
Read the full abstract

Computation and Language

Language as a Latent Variable for Reasoning Optimization

Linjuan Wu, Haoran Wei, Jialong Tang et al. (+4 more)

As LLMs reduce English-centric bias, a surprising trend emerges: non-English responses sometimes outperform English on reasoning tasks. We hypothesize that language functions as a latent variable that structurally modulates the model's internal inference pathways, rather than merely serving as an output medium. To test this, we conducted a Polyglot Thinking Experiment, in which models were prompted to solve identical problems under language-constrained and language-unconstrained conditions. Results show that non-English responses often achieve higher accuracy, and the best performance frequently occur when language is unconstrained, suggesting that multilinguality broadens the model's latent reasoning space. Based on this insight, we propose polyGRPO (Polyglot Group Relative Policy Optimization), an RL framework that treats language variation as an implicit exploration signal. It generates polyglot preference data online under language-constrained and unconstrained conditions, optimizing the policy with respect to both answer accuracy and reasoning structure. Trained on only 18.1K multilingual math problems without chain-of-thought annotations, polyGRPO improves the base model (Qwen2.5-7B-Instruct) by 6.72% absolute accuracy on four English reasoning testset and 6.89% in their multilingual benchmark. Remarkably, it is the only method that surpasses the base LLM on English commonsense reasoning task (4.9%), despite being trained solely on math data-highlighting its strong cross-task generalization. Further analysis reveals that treating language as a latent variable expands the model's latent reasoning space, yielding consistent and generalizable improvements in reasoning performance.
Read the full abstract

All authors Linjuan Wu, Haoran Wei, Jialong Tang, Shuang Luo, Baosong Yang, Yongliang Shen, Weiming Lu

Computation and Language

AgenticQwen: Training Small Agentic Language Models with Dual Data Flywheels for Industrial-Scale Tool Use

Yuanjie Lyu, Chengyu Wang, Haonan Zheng et al. (+4 more)

Modern industrial applications increasingly demand language models that act as agents, capable of multi-step reasoning and tool use in real-world settings. These tasks are typically performed under strict cost and latency constraints, making small agentic models highly desirable. In this paper, we introduce the AgenticQwen family of models, trained via multi-round reinforcement learning (RL) on synthetic data and a limited amount of open-source data. Our training framework combines reasoning RL and agentic RL with dual data flywheels that automatically generate increasingly challenging tasks. The reasoning flywheel increases task difficulty by learning from errors, while the agentic flywheel expands linear workflows into multi-branch behavior trees that better reflect the decision complexity of real-world applications. We validate AgenticQwen on public benchmarks and in an industrial agent system. The models achieve strong performance on multiple agentic benchmarks, and in our industrial agent system, close the gap with much larger models on search and data analysis tasks. Model checkpoints and part of the synthetic data: https://huggingface.co/collections/alibaba-pai/agenticqwen. Data synthesis and RL training code: https://github.com/haruhi-sudo/data_synth_and_rl. The data synthesis pipeline is also integrated into EasyDistill: https://github.com/modelscope/easydistill.
Read the full abstract

All authors Yuanjie Lyu, Chengyu Wang, Haonan Zheng, Yuanhao Yue, Junbing Yan, Ming Wang, Jun Huang

Computation and Language

Measuring Opinion Bias and Sycophancy via LLM-based Persuasion

Rodrigo Nogueira, Giovana Kerche Bonás, Thales Sales Almeida et al. (+7 more)

Large language models increasingly shape the information people consume: they are embedded in search, consulted for professional advice, deployed as agents, and used as a first stop for questions about policy, ethics, health, and politics. When such a model silently holds a position on a contested topic, that position propagates at scale into users' decisions. Eliciting a model's positions is harder than it first appears: contemporary assistants answer direct opinion questions with evasive disclaimers, and the same model may concede the opposite position once the user starts arguing one side. We propose a method, released as the open-source llm-bias-bench, for discovering the opinions an LLM actually holds on contested topics under conditions that resemble real multi-turn interaction. The method pairs two complementary free-form probes. Direct probing asks for the model's opinion across five turns of escalating pressure from a simulated user. Indirect probing never asks for an opinion and engages the model in argumentative debate, letting bias leak through how it concedes, resists, or counter-argues. Three user personas (neutral, agree, disagree) collapse into a nine-way behavioral classification that separates persona-independent positions from persona-dependent sycophancy, and an auditable LLM judge produces verdicts with textual evidence. The first instantiation ships 38 topics in Brazilian Portuguese across values, scientific consensus, philosophy, and economic policy. Applied to 13 assistants, the method surfaces findings of practical interest: argumentative debate triggers sycophancy 2-3x more than direct questioning (median 50% to 79%); models that look opinionated under direct questioning often collapse into mirroring under sustained arguments; and attacker capability matters mainly when an existing opinion must be dislodged, not when the assistant starts neutral.
Read the full abstract

All authors Rodrigo Nogueira, Giovana Kerche Bonás, Thales Sales Almeida, Andrea Roque, Ramon Pires, Hugo Abonizio, Thiago Laitz, Celio Larcher, Roseval Malaquias Junior, Marcos Piau

Computation and Language

Finding Meaning in Embeddings: Concept Separation Curves

Paul Keuren, Marc Ponsen and Robert Ayoub Bagheri

Sentence embedding techniques aim to encode key concepts of a sentence's meaning in a vector space. However, the majority of evaluation approaches for sentence embedding quality rely on the use of additional classifiers or downstream tasks. These additional components make it unclear whether good results stem from the embedding itself or from the classifier's behaviour. In this paper, we propose a novel method for evaluating the effectiveness of sentence embedding methods in capturing sentence-level concepts. Our approach is classifier-independent, allowing for an objective assessment of the model's performance. The approach adopted in this study involves the systematic introduction of syntactic noise and semantic negations into sentences, with the subsequent quantification of their relative effects on the resulting embeddings. The visualisation of these effects is facilitated by Concept Separation Curves, which show the model's capacity to differentiate between conceptual and surface-level variations. By leveraging data from multiple domains, employing both Dutch and English languages, and examining sentence lengths, this study offers a compelling demonstration that Concept Separation Curves provide an interpretable, reproducible, and cross-model approach for evaluating the conceptual stability of sentence embeddings.
Read the full abstract

Computation and Language

UKP_Psycontrol at SemEval-2026 Task 2: Modeling Valence and Arousal Dynamics from Text

Darya Hryhoryeva, Amaia Zurinaga, Hamidreza Jamalabadi et al. (+1 more)

This paper presents our system developed for SemEval-2026 Task 2. The task requires modeling both current affect and short-term affective change in chronologically ordered user-generated texts. We explore three complementary approaches: (1) LLM prompting under user-aware and user-agnostic settings, (2) a pairwise Maximum Entropy (MaxEnt) model with Ising-style interactions for structured transition modeling, and (3) a lightweight neural regression model incorporating recent affective trajectories and trainable user embeddings. Our findings indicate that LLMs effectively capture static affective signals from text, whereas short-term affective variation in this dataset is more strongly explained by recent numeric state trajectories than by textual semantics. Our system ranked first among participating teams in both Subtask 1 and Subtask 2A based on the official evaluation metric.
Read the full abstract

All authors Darya Hryhoryeva, Amaia Zurinaga, Hamidreza Jamalabadi, Iryna Gurevych

Computation and Language

Job Skill Extraction via LLM-Centric Multi-Module Framework

Guojing Li, Zichuan Fu, Junyi Li et al. (+8 more)

Span-level skill extraction from job advertisements underpins candidate-job matching and labor-market analytics, yet generative large language models (LLMs) often yield malformed spans, boundary drift, and hallucinations, especially with long-tail terms and cross-domain shift. We present SRICL, an LLM-centric framework that combines semantic retrieval (SR), in-context learning (ICL), and supervised fine-tuning (SFT) with a deterministic verifier. SR pulls in-domain annotated sentences and definitions from ESCO to form format-constrained prompts that stabilize boundaries and handle coordination. SFT aligns output behavior, while the verifier enforces pairing, non-overlap, and BIO legality with minimal retries. On six public span-labeled corpora of job-ad sentences across sectors and languages, SRICL achieves substantial STRICT-F1 improvements over GPT-3.5 prompting baselines and sharply reduces invalid tags and hallucinated spans, enabling dependable sentence-level deployment in low-resource, multi-domain settings.
Read the full abstract

All authors Guojing Li, Zichuan Fu, Junyi Li, Faxue Liu, Wenxia Zhou, Yejing Wang, Jingtong Gao, Maolin Wang, Rungen Liu, Wenlin Zhang, Xiangyu Zhao

Computation and Language · Cross-listed from cs.CV

Seeing Isn't Believing: Uncovering Blind Spots in Evaluator Vision-Language Models

Mohammed Safi Ur Rahman Khan, Sanjay Suryanarayanan, Tushar Anand et al. (+1 more)

Large Vision-Language Models (VLMs) are increasingly used to evaluate outputs of other models, for image-to-text (I2T) tasks such as visual question answering, and text-to-image (T2I) generation tasks. Despite this growing reliance, the reliability of these Evaluator VLMs remains under explored. In this work, we systematically evaluate the reliability of Evaluator VLMs across both I2T and T2I tasks. We introduce targeted perturbations that degrade output quality along key error dimensions, including object hallucinations, spatial reasoning, factual grounding, and visual fidelity. These perturbations test whether Evaluator VLMs can reliably account for these quality degrading errors in their evaluations. Using a comprehensive benchmark of over 4000 perturbed instances spanning 40 perturbation dimensions, we evaluate 4 prominent VLMs using single-answer scoring, pairwise comparison, and reference-guided paradigms. Our findings reveal that current VLM evaluators exhibit substantial blind spots: they often fail to detect perturbed outputs - in some cases exceeding 50%, struggle particularly with fine-grained compositional and spatial errors, and are often insensitive to hallucinated content that contradicts the input image. Pairwise comparison proves more reliable, though failure rates persist. These results highlight the unreliable nature of current Evaluator VLMs and urge caution in their deployment for benchmarking and development decisions. Code and data have been made publicly available.
Read the full abstract

All authors Mohammed Safi Ur Rahman Khan, Sanjay Suryanarayanan, Tushar Anand, Mitesh M. Khapra

Computation and Language · Cross-listed from cs.IR

From Tokens to Concepts: Leveraging SAE for SPLADE

Yuxuan Zong, Mathias Vast, Basile Van Cooten et al. (+2 more)

Learned Sparse IR models, such as SPLADE, offer an excellent efficiency-effectiveness tradeoff. However, they rely on the underlying backbone vocabulary, which might hinder performance (polysemicity and synonymy) and pose a challenge for multi-lingual and multi-modal usages. To solve this limitation, we propose to replace the backbone vocabulary with a latent space of semantic concepts learned using Sparse Auto-Encoders (SAE). Throughout this paper, we study the compatibility of these 2 concepts, explore training approaches, and analyze the differences between our SAE-SPLADE model and traditional SPLADE models. Our experiments demonstrate that SAE-SPLADE achieves retrieval performance comparable to SPLADE on both in-domain and out-of-domain tasks while offering improved efficiency.
Read the full abstract

All authors Yuxuan Zong, Mathias Vast, Basile Van Cooten, Laure Soulier, Benjamin Piwowarski

Computation and Language

OptiVerse: A Comprehensive Benchmark towards Optimization Problem Solving

Xinyu Zhang, Boxuan Zhang, Yuchen Wan et al. (+5 more)

While Large Language Models (LLMs) demonstrate remarkable reasoning, complex optimization tasks remain challenging, requiring domain knowledge and robust implementation. However, existing benchmarks focus narrowly on Mathematical Programming and Combinatorial Optimization, hindering comprehensive evaluation. To address this, we introduce OptiVerse, a comprehensive benchmark of 1,000 curated problems spanning neglected domains, including Stochastic Optimization, Dynamic Optimization, Game Optimization, and Optimal Control, across three difficulty levels: Easy, Medium, and Hard. The experiments with 22 LLMs of different sizes reveal sharp performance degradation on hard problems, where even advanced models like GPT-5.2 and Gemini-3 struggle to exceed 27% accuracy. Through error analysis, we identify that modeling & logic errors remain the primary bottleneck. Consequently, we propose a Dual-View Auditor Agent that improves the accuracy of the LLM modeling process without introducing significant time overhead. OptiVerse will serve as a foundational platform for advancing LLMs in solving complex optimization challenges.
Read the full abstract

All authors Xinyu Zhang, Boxuan Zhang, Yuchen Wan, Lingling Zhang, YiXing Yao, Bifan Wei, Yaqiang Wu, Jun Liu

Computation and Language · Cross-listed from cs.AI

How English Print Media Frames Human-Elephant Conflicts in India

Bonala Sai Punith, Salveru Jayati, Garima Shakya et al. (+1 more)

Human-elephant conflict (HEC) is rising across India as habitat loss and expanding human settlements force elephants into closer contact with people. While the ecological drivers of conflict are well-studied, how the news media portrays them remains largely unexplored. This work presents the first large-scale computational analysis of media framing of HEC in India, examining 1,968 full-length news articles consisting of 28,986 sentences, from a major English-language outlet published between January 2022 and September 2025. Using a multi-model sentiment framework that combines long-context transformers, large language models, and a domain-specific Negative Elephant Portrayal Lexicon, we quantify sentiment, extract rationale sentences, and identify linguistic patterns that contribute to negative portrayals of elephants. Our findings reveal a dominance of fear-inducing and aggression-related language. Since the media framing can shape public attitudes toward wildlife and conservation policy, such narratives risk reinforcing public hostility and undermining coexistence efforts. By providing a transparent, scalable methodology and releasing all resources through an anonymized repository, this study highlights how Web-scale text analysis can support responsible wildlife reporting and promote socially beneficial media practices.
Read the full abstract

All authors Bonala Sai Punith, Salveru Jayati, Garima Shakya, Shubham Kumar Nigam