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

Friday 9 January

Computation and Language

Pantagruel: Unified Self-Supervised Encoders for French Text and Speech

Phuong-Hang Le, Valentin Pelloin, Arnault Chatelain et al. (+27 more)

We release Pantagruel models, a new family of self-supervised encoder models for French text and speech. Instead of predicting modality-tailored targets such as textual tokens or speech units, Pantagruel learns contextualized target representations in the feature space, allowing modality-specific encoders to capture linguistic and acoustic regularities more effectively. Separate models are pre-trained on large-scale French corpora, including Wikipedia, OSCAR and CroissantLLM for text, together with MultilingualLibriSpeech, LeBenchmark, and INA-100k for speech. INA-100k is a newly introduced 100,000-hour corpus of French audio derived from the archives of the Institut National de l'Audiovisuel (INA), the national repository of French radio and television broadcasts, providing highly diverse audio data. We evaluate Pantagruel across a broad range of downstream tasks spanning both modalities, including those from the standard French benchmarks such as FLUE or LeBenchmark. Across these tasks, Pantagruel models show competitive or superior performance compared to strong French baselines such as CamemBERT, FlauBERT, and LeBenchmark2.0, while maintaining a shared architecture that can seamlessly handle either speech or text inputs. These results confirm the effectiveness of feature-space self-supervised objectives for French representation learning and highlight Pantagruel as a robust foundation for multimodal speech-text understanding.
Read the full abstract

All authors Phuong-Hang Le, Valentin Pelloin, Arnault Chatelain, Maryem Bouziane, Mohammed Ghennai, Qianwen Guan, Kirill Milintsevich, Salima Mdhaffar, Aidan Mannion, Nils Defauw, Shuyue Gu, Alexandre Audibert, Marco Dinarelli, Yannick Estève, Lorraine Goeuriot, Steffen Lalande, Nicolas Hervé, Maximin Coavoux, François Portet, Étienne Ollion, Marie Candito, Maxime Peyrard, Solange Rossato, Benjamin Lecouteux, Aurélie Nardy, Gilles Sérasset, Vincent Segonne, Solène Evain, Diandra Fabre, Didier Schwab

Computation and Language

Illusions of Confidence? Diagnosing LLM Truthfulness via Neighborhood Consistency

Haoming Xu, Ningyuan Zhao, Yunzhi Yao et al. (+7 more)

As Large Language Models (LLMs) are increasingly deployed in real-world settings, correctness alone is insufficient. Reliable deployment requires maintaining truthful beliefs under contextual perturbations. Existing evaluations largely rely on point-wise confidence like Self-Consistency, which can mask brittle belief. We show that even facts answered with perfect self-consistency can rapidly collapse under mild contextual interference. To address this gap, we propose Neighbor-Consistency Belief (NCB), a structural measure of belief robustness that evaluates response coherence across a conceptual neighborhood. To validate the efficiency of NCB, we introduce a new cognitive stress-testing protocol that probes outputs stability under contextual interference. Experiments across multiple LLMs show that the performance of high-NCB data is relatively more resistant to interference. Finally, we present Structure-Aware Training (SAT), which optimizes context-invariant belief structure and reduces long-tail knowledge brittleness by approximately 30%. Code is available at https://github.com/zjunlp/belief.
Read the full abstract

All authors Haoming Xu, Ningyuan Zhao, Yunzhi Yao, Weihong Xu, Hongru Wang, Xinle Deng, Shumin Deng, Jeff Z. Pan, Huajun Chen, Ningyu Zhang

Computation and Language

HAPS: Hierarchical LLM Routing with Joint Architecture and Parameter Search

Zihang Tian, Rui Li, Jingsen Zhang et al. (+3 more)

Large language model (LLM) routing aims to exploit the specialized strengths of different LLMs for diverse tasks. However, existing approaches typically focus only on selecting LLM architectures, treating candidate models as fixed while overlooking parameter adaptation, which can substantially affect task performance. In this paper, we introduce HAPS, a hierarchical LLM routing framework that jointly searches over model architectures and parameters. Specifically, a high-level router selects among candidate LLM architectures, while a low-level router generates input-conditioned LoRA parameters for the selected architecture, expanding each candidate from a single fixed behavior into a region of reachable behaviors. The low-level router is built on the backbone of the high-level router, so that architecture selection and parameter generation share a common representation of the prompt. We further optimize the whole framework with a reinforcement learning objective. Extensive experiments show that HAPS consistently outperforms strong routing baselines across reasoning- and knowledge-intensive benchmarks, with analyses of its components, design choices, scalability, cost, and mixed open- and closed-source deployment further demonstrating the robustness of these gains. Our code is available at https://github.com/zihangtian/HAPS.
Read the full abstract

All authors Zihang Tian, Rui Li, Jingsen Zhang, Xiaohe Bo, Wei Huo, Xu Chen

Computation and Language

An Empirical Study on Preference Tuning Generalization and Diversity Under Domain Shift

Constantinos Karouzos, Xingwei Tan and Nikolaos Aletras

Preference tuning aligns base language models to human judgments of quality, helpfulness, or safety by optimizing over explicit preference signals rather than likelihood alone. Prior work has shown that preference tuning degrades performance and reduces helpfulness outside the training domain. However, the extent to which adaptation strategies mitigate this domain shift remains unexplored. We address this challenge by conducting a comprehensive and systematic study of alignment generalization under domain shift. We compare five popular alignment objectives and various adaptation strategies from source to target, including target-domain supervised fine-tuning and pseudo-labeling, across summarization, question-answering helpfulness, and safety alignment tasks. Our findings reveal systematic differences in generalization across alignment objectives under domain shift. We show that adaptation strategies based on pseudo-labeling substantially reduce domain-shift degradation but induce mode collapse, revealing a generalization-diversity trade-off.
Read the full abstract

Computation and Language

Gender Bias in LLMs: Preliminary Evidence from Shared Parenting Scenario in Czech Family Law

Jakub Harasta, Matej Vasina, Martin Kornel et al. (+1 more)

Access to justice remains limited for many people, leading laypersons to increasingly rely on Large Language Models (LLMs) for legal self-help. Laypeople use these tools intuitively, which may lead them to form expectations based on incomplete, incorrect, or biased outputs. This study examines whether leading LLMs exhibit gender bias in their responses to a realistic family law scenario. We present an expert-designed divorce scenario grounded in Czech family law and evaluate four state-of-the-art LLMs GPT-5 nano, Claude Haiku 4.5, Gemini 2.5 Flash, and Llama 3.3 in a fully zero-shot interaction. We deploy two versions of the scenario, one with gendered names and one with neutral labels, to establish a baseline for comparison. We further introduce nine legally relevant factors that vary the factual circumstances of the case and test whether these variations influence the models' proposed shared-parenting ratios. Our preliminary results highlight differences across models and suggest gender-dependent patterns in the outcomes generated by some systems. The findings underscore both the risks associated with laypeople's reliance on LLMs for legal guidance and the need for more robust evaluation of model behavior in sensitive legal contexts. We present exploratory and descriptive evidence intended to identify systematic asymmetries rather than to establish causal effects.
Read the full abstract

All authors Jakub Harasta, Matej Vasina, Martin Kornel, Tomas Foltynek

Computation and Language

iReasoner: Trajectory-Aware Intrinsic Reasoning Supervision for Self-Evolving Large Multimodal Models

Meghana Sunil, Manikandarajan Venmathimaran and Muthu Subash Kavitha

Recent work shows that large multimodal models (LMMs) can self-improve from unlabeled data via self-play and intrinsic feedback. Yet existing self-evolving frameworks mainly reward final outcomes, leaving intermediate reasoning weakly constrained despite its importance for visually grounded decision making. We propose iReasoner, a self-evolving framework that improves an LMM's implicit reasoning by explicitly eliciting chain-of-thought (CoT) and rewarding its internal agreement. In a Proposer–Solver loop over unlabeled images, iReasoner augments outcome-level intrinsic rewards with a trajectory-aware signal defined over intermediate reasoning steps, providing learning signals that distinguish reasoning paths leading to the same answer without ground-truth labels or external judges. Starting from Qwen2.5-VL-7B, iReasoner yields up to $+2.1$ points across diverse multimodal reasoning benchmarks under fully unsupervised post-training. We hope this work serves as a starting point for reasoning-aware self-improvement in LMMs in purely unsupervised settings. Our code is available at https://meghanaasunil.github.io/iReasoner.
Read the full abstract

Computation and Language

Continual-learning for Modelling Low-Resource Languages from Large Language Models

Santosh Srinath K, Mudit Somani, Varun Reddy Padala et al. (+2 more)

Modelling a language model for a multi-lingual scenario includes several potential challenges, among which catastrophic forgetting is the major challenge. For example, small language models (SLM) built for low-resource languages by adapting large language models (LLMs) pose the challenge of catastrophic forgetting. This work proposes to employ a continual learning strategy using parts-of-speech (POS)-based code-switching along with a replay adapter strategy to mitigate the identified gap of catastrophic forgetting while training SLM from LLM. Experiments conducted on vision language tasks such as visual question answering and language modelling task exhibits the success of the proposed architecture.
Read the full abstract

All authors Santosh Srinath K, Mudit Somani, Varun Reddy Padala, Prajna Devi Upadhyay, Abhijit Das

Computation and Language

FACTUM: Mechanistic Detection of Citation Hallucination in Long-Form RAG

Maxime Dassen, Rebecca Kotula, Kenton Murray et al. (+5 more)

Retrieval-Augmented Generation (RAG) models are critically undermined by citation hallucinations, a deceptive failure where a model cites a source that fails to support its claim. While existing work attributes hallucination to a simple over-reliance on parametric knowledge, we reframe this failure as an evolving, scale-dependent coordination failure between the Attention (reading) and Feed-Forward Network (recalling) pathways. We introduce FACTUM (Framework for Attesting Citation Trustworthiness via Underlying Mechanisms), a framework of four mechanistic scores: Contextual Alignment (CAS), Attention Sink Usage (BAS), Parametric Force (PFS), and Pathway Alignment (PAS). Our analysis reveals that correct citations are consistently marked by higher parametric force (PFS) and greater use of the attention sink (BAS) for information synthesis. Crucially, we find that "one-size-fits-all" theories are insufficient as the signature of correctness evolves with scale: while the 3B model relies on high pathway alignment (PAS), our best-performing 8B detector identifies a shift toward a specialized strategy where pathways provide distinct, orthogonal information. By capturing this complex interplay, FACTUM outperforms state-of-the-art baselines by up to 37.5% in AUC. Our results demonstrate that high parametric force is constructive when successfully coordinated with the Attention pathway, paving the way for more nuanced and reliable RAG systems.
Read the full abstract

All authors Maxime Dassen, Rebecca Kotula, Kenton Murray, Andrew Yates, Dawn Lawrie, Efsun Kayi, James Mayfield, Kevin Duh

Computation and Language

What do the metrics mean? A critical analysis of the use of Automated Evaluation Metrics in Interpreting

Jonathan Downie and Joss Moorkens

With the growth of interpreting technologies, from remote interpreting and Computer-Aided Interpreting to automated speech translation and interpreting avatars, there is now a high demand for ways to quickly and efficiently measure the quality of any interpreting delivered. A range of approaches to fulfil the need for quick and efficient quality measurement have been proposed, each involving some measure of automation. This article examines these recently-proposed quality measurement methods and will discuss their suitability for measuring the quality of authentic interpreting practice, whether delivered by humans or machines, concluding that automatic metrics as currently proposed cannot take into account the communicative context and thus are not viable measures of the quality of any interpreting provision when used on their own. Across all attempts to measure or even categorise quality in Interpreting Studies, the contexts in which interpreting takes place have become fundamental to the final analysis.
Read the full abstract

Computation and Language

CLewR: Curriculum Learning with Restarts for Machine Translation Preference Learning

Alexandra Dragomir, Florin Brad and Radu Tudor Ionescu

Large language models (LLMs) have demonstrated competitive performance in zero-shot multilingual machine translation (MT). Some follow-up works further improved MT performance via preference optimization, but they leave a key aspect largely underexplored: the order in which data samples are given during training. We address this topic by integrating curriculum learning into various state-of-the-art preference optimization algorithms to boost MT performance. We introduce a novel curriculum learning strategy with restarts (CLewR), which reiterates easy-to-hard curriculum multiple times during training to effectively mitigate the catastrophic forgetting of easy examples. We demonstrate consistent gains across several model families (Gemma2, Qwen2.5, Llama3.1) and preference optimization techniques. We publicly release our code at https://github.com/alexandra-dragomir/CLewR.
Read the full abstract

Computation and Language

Router-Suggest: Dynamic Routing for Multimodal Auto-Completion in Visually-Grounded Dialogs

Sandeep Mishra, Devichand Budagam, Anubhab Mandal et al. (+3 more)

Real-time multimodal auto-completion is essential for digital assistants, chatbots, design tools, and healthcare consultations, where user inputs rely on shared visual context. We introduce Multimodal Auto-Completion (MAC), a task that predicts upcoming characters in live chats using partially typed text and visual cues. Unlike traditional text-only auto-completion (TAC), MAC grounds predictions in multimodal context to better capture user intent. To enable this task, we adapt MMDialog and ImageChat to create benchmark datasets. We evaluate leading vision-language models (VLMs) against strong textual baselines, highlighting trade-offs in accuracy and efficiency. We present Router-Suggest, a router framework that dynamically selects between textual models and VLMs based on dialog context, along with a lightweight variant for resource-constrained environments. Router-Suggest achieves a 2.3x to 10x speedup over the best-performing VLM. A user study shows that VLMs significantly excel over textual models on user satisfaction, notably saving user typing effort and improving the quality of completions in multi-turn conversations. These findings underscore the need for multimodal context in auto-completions, leading to smarter, user-aware assistants.
Read the full abstract

All authors Sandeep Mishra, Devichand Budagam, Anubhab Mandal, Bishal Santra, Pawan Goyal, Manish Gupta

Computation and Language

Schema-Grounded LLM Extraction for FHIR Patient Digital Twins

Rafael Brens, Yuqiao Meng, Luoxi Tang et al. (+1 more)

We revisit the problem of constructing interoperable patient digital twins from unstructured electronic health records (EHRs) and argue that the task is better cast not as a cascade of extraction modules but as constrained generation of a valid FHIR bundle. We introduce SG-LLM, a schema-grounded LLM extractor that (i) augments the prompt with candidate SNOMED-CT, RxNorm, and LOINC codes retrieved through a SapBERT index, (ii) decodes under a JSON Schema derived directly from FHIR R4 StructureDefinitions, and (iii) closes a validator-in-the-loop repair stage whose diagnostics are fed back as structured error messages. We argue that the twin's usefulness, not only span-level F1, is the right object of evaluation, and operationalize this with a clinical-utility experiment that measures the gap in 30-day readmission AUROC between classifiers trained on SG-LLM-generated FHIR bundles versus expert-curated ones. On MIMIC-IV and n2c2 2018 Track 2 benchmarks, SG-LLM matches or exceeds strong joint-extraction and vanilla-LLM baselines while producing substantially more valid bundles. Ablations isolate the contributions of retrieval, schema constraint, and the repair loop. All code, prompts, and schemas are released.
Read the full abstract

All authors Rafael Brens, Yuqiao Meng, Luoxi Tang, Zhaohan Xi

Computation and Language

Left, Right, or Center? Evaluating LLM Framing in News Classification and Generation

Molly Kennedy, Ali Parker, Yihong Liu et al. (+1 more)

Large Language Model (LLM) based summarization and text generation are increasingly used for producing and rewriting text, raising concerns about political framing in journalism where subtle wording choices can shape interpretation. Across nine state-of-the-art LLMs, we study political framing by testing whether LLMs' classification-based bias signals align with framing behavior in their generated summaries. We first compare few-shot ideology predictions against LEFT/CENTER/RIGHT labels. We then generate "steered" summaries under FAITHFUL, CENTRIST, LEFT, and RIGHT prompts, and score all outputs using a single fixed ideology evaluator. We find pervasive ideological center-collapse in both article-level ratings and generated text, indicating a systematic tendency toward centrist framing. Among evaluated models, Grok 4 is by far the most ideologically expressive generator, while Claude Sonnet 4.5 and Llama 3.1 achieve the strongest bias-rating performance among commercial and open-weight models, respectively.
Read the full abstract

All authors Molly Kennedy, Ali Parker, Yihong Liu, Hinrich Schütze

Computation and Language

Peek2: Regex-free Byte-level Byte-Pair Encoding Pretokenizer for LLM Inference on Edge Devices

Liu Zai and Iraklis Klampanos

Pretokenization is a crucial, sequential pass in Byte-level BPE tokenizers, yet little work has been done to optimize it for edge-side inference. Our proposed new implementation, Peek2, serves as a drop-in replacement for cl100k-like pretokenizers used in GPT-3, LLaMa-3, and Qwen-2.5. After breaking down and analyzing the logic of the original cl100k pretokenizer, we introduced a new pretokenization algorithm with linear time complexity and constant, trivial memory usage, suited for edge scenarios. Test results show that it increases microbenchmarking throughput by up to $ 2.48\times $ and delivers a $ 1.14\times $ improvement in overall throughput across the entire Byte-level BPE encoding process, depending on the dataset, while providing identical results as the baseline Regex-based tokenizer.
Read the full abstract

Computation and Language

LLMs as Science Journalists: Supporting Early-stage Researchers in Communicating Their Science to the Public

Milad Alshomary, Grace Li, Anubhav Jangra et al. (+3 more)

The scientific community needs tools that help early-stage researchers effectively communicate their findings and innovations to the public. Although existing general-purpose Large Language Models (LLMs) can assist in this endeavor, they are not optimally aligned for it. To address this, we propose a framework for training LLMs to emulate the role of a science journalist that can be used by early-stage researchers to learn how to properly communicate their papers to the general public. We evaluate the usefulness of our trained LLM Journalists in leading conversations with both simulated and human researchers. %compared to the general-purpose ones. Our experiments indicate that LLMs trained using our framework ask more relevant questions that address the societal impact of research, prompting researchers to clarify and elaborate on their findings. In the user study, the majority of participants who interacted with our trained LLM Journalist appreciated it more than interacting with general-purpose LLMs.
Read the full abstract

All authors Milad Alshomary, Grace Li, Anubhav Jangra, Yufang Hou, Kathleen McKeown, Smaranda Muresan

Computation and Language

EnvScaler: Scaling Tool-Interactive Environments for LLM Agent via Programmatic Synthesis

Xiaoshuai Song, Haofei Chang, Guanting Dong et al. (+3 more)

Large language models (LLMs) are expected to be trained to act as agents in various real-world environments, but this process relies on rich and varied tool-interaction sandboxes. However, access to real systems is often restricted; LLM-simulated environments are prone to hallucinations and inconsistencies; and manually built sandboxes are hard to scale. In this paper, we propose EnvScaler, an automated framework for scalable tool-interaction environments via programmatic synthesis. EnvScaler comprises two components. First, SkelBuilder constructs diverse environment skeletons through topic mining, logic modeling, and quality evaluation. Then, ScenGenerator generates multiple task scenarios and rule-based trajectory validation functions for each environment. With EnvScaler, we synthesize 191 environments and about 7K scenarios, and apply them to Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL) for Qwen3 series models. Results on three benchmarks show that EnvScaler significantly improves LLMs' ability to solve tasks in complex environments involving multi-turn, multi-tool interactions. We release our code and data at https://github.com/RUC-NLPIR/EnvScaler.
Read the full abstract

All authors Xiaoshuai Song, Haofei Chang, Guanting Dong, Yutao Zhu, Ji-Rong Wen, Zhicheng Dou

Computation and Language · Cross-listed from cs.LG

Fusion Matters: Length-Aware Analysis of Positional-Encoding Fusion in Transformers

Mohamed Amine Hallam and Kuo-Kun Tseng

Transformers require positional encodings to represent sequence order, yet most prior work focuses on designing new positional encodings rather than examining how positional information is fused with token embeddings. In this paper, we study whether the fusion mechanism itself affects performance, particularly in long-sequence settings. We conduct a controlled empirical study comparing three canonical fusion strategies–element-wise addition, concatenation with projection, and scalar gated fusion–under identical Transformer architectures, data splits, and random seeds. Experiments on three text classification datasets spanning short (AG News), medium (IMDB), and long (ArXiv) sequences show that fusion choice has negligible impact on short texts but produces consistent gains on long documents. To verify that these gains are structural rather than stochastic, we perform paired-seed analysis and cross-dataset comparison across sequence-length regimes. Additional experiments on the ArXiv dataset indicate that the benefit of learnable fusion generalizes across multiple positional encoding families. Finally, we explore a lightweight convolutional gating mechanism that introduces local inductive bias at the fusion level, evaluated on long documents only. Our results indicate that positional-encoding fusion is a non-trivial design choice for long-sequence Transformers and should be treated as an explicit modeling decision rather than a fixed default.
Read the full abstract

Computation and Language

Simplify-This: A Comparative Analysis of Prompt-Based and Fine-Tuned LLMs

Eilam Cohen, Itamar Bul, Danielle Inbar et al. (+1 more)

Large language models (LLMs) enable strong text generation, and in general there is a practical tradeoff between fine-tuning and prompt engineering. We introduce Simplify-This, a comparative study evaluating both paradigms for text simplification with encoder-decoder LLMs across multiple benchmarks, using a range of evaluation metrics. Fine-tuned models consistently deliver stronger structural simplification, whereas prompting often attains higher semantic similarity scores yet tends to copy inputs. A human evaluation favors fine-tuned outputs overall. We release code, a cleaned derivative dataset used in our study, checkpoints of fine-tuned models, and prompt templates to facilitate reproducibility and future work.
Read the full abstract

All authors Eilam Cohen, Itamar Bul, Danielle Inbar, Omri Loewenbach

Computation and Language

One Script Instead of Hundreds? On Pretraining Romanized Encoder Language Models

Benedikt Ebing, Lennart Keller and Goran Glavaš

Exposing latent lexical overlap, script romanization has emerged as an effective strategy for improving cross-lingual transfer (XLT) in multilingual language models (mLMs). Most prior work, however, focused on setups that favor romanization the most: (1) transfer from high-resource Latin-script to low-resource non-Latin-script languages and/or (2) between genealogically closely related languages with different scripts. It thus remains unclear whether romanization is a good representation choice for pretraining general-purpose mLMs, or, more precisely, if information loss associated with romanization harms performance for high-resource languages. We address this gap by pretraining encoder LMs from scratch on both romanized and original texts for six typologically diverse high-resource languages, investigating two potential sources of degradation: (i) loss of script-specific information and (ii) negative cross-lingual interference from increased vocabulary overlap. Using two romanizers with different fidelity profiles, we observe negligible performance loss for languages with segmental scripts, whereas languages with morphosyllabic scripts (Chinese and Japanese) suffer degradation that higher-fidelity romanization mitigates but cannot fully recover. Importantly, comparing monolingual LMs with their mLM counterpart, we find no evidence that increased subword overlap induces negative interference. We further show that romanization improves encoding efficiency (i.e., fertility) for segmental scripts at a negligible performance cost.
Read the full abstract

Computation and Language · Cross-listed from cs.LG

Weights to Code: Extracting Interpretable Algorithms from the Discrete Transformer

Yifan Zhang, Wei Bi, Kechi Zhang et al. (+3 more)

Algorithm extraction aims to synthesize executable programs directly from models trained on algorithmic tasks, enabling de novo recovery of executable mechanisms from weights without relying on human-written target programs. However, applying this paradigm to Transformer is complicated by representation entanglement (e.g., superposition), where features encoded in overlapping directions substantially hinder the recovery of symbolic expressions. We propose the Discrete Transformer, an architecture explicitly designed to bridge the gap between continuous representations and discrete symbolic logic. By injecting discreteness through temperature-annealed sampling, our framework effectively leverages hypothesis testing and symbolic regression to extract human-readable programs. Empirically, the Discrete Transformer achieves performance comparable to the RNN-based MIPS baseline on shared discrete tasks, while broadening extraction to tasks with continuous-valued intermediate computations. Finally, we show that architectural inductive biases provide fine-grained control over synthesized programs, establishing the Discrete Transformer as a controllable testbed for algorithm extraction and Transformer interpretability.
Read the full abstract

All authors Yifan Zhang, Wei Bi, Kechi Zhang, Dongming Jin, Jie Fu, Zhi Jin

Computation and Language

AutoMonitor-Bench: Evaluating the Reliability of LLM-Based Misbehavior Monitor

Shu Yang, Jingyu Hu, Tong Li et al. (+3 more)

We introduce AutoMonitor-Bench, the first benchmark designed to systematically evaluate the reliability of LLM-based misbehavior monitors across diverse tasks and failure modes. AutoMonitor-Bench consists of 3,010 carefully annotated test samples spanning question answering, code generation, and reasoning, with paired misbehavior and benign instances. We evaluate monitors using two complementary metrics: Miss Rate (MR) and False Alarm Rate (FAR), capturing failures to detect misbehavior and oversensitivity to benign behavior, respectively. Evaluating 12 proprietary and 10 open-source LLMs, we observe substantial variability in monitoring performance and a consistent trade-off between MR and FAR, revealing an inherent safety-utility tension. To further explore the limits of monitor reliability, we construct a large-scale training corpus of 153,581 samples and fine-tune Qwen3-4B-Instruction to investigate whether training on known, relatively easy-to-construct misbehavior datasets improves monitoring performance on unseen and more implicit misbehaviors. Our results highlight the challenges of reliable, scalable misbehavior monitoring and motivate future work on task-aware designing and training strategies for LLM-based monitors.
Read the full abstract

All authors Shu Yang, Jingyu Hu, Tong Li, Hanqi Yan, Wenxuan Wang, Di Wang

Computation and Language

Analysing Differences in Persuasive Language in LLM-Generated Text: Uncovering Stereotypical Gender Patterns

Amalie Brogaard Pauli, Maria Barrett, Max Müller-Eberstein et al. (+2 more)

Large language models (LLMs) are increasingly used for everyday communication tasks, including drafting interpersonal messages intended to influence and persuade. Prior work has shown that LLMs can successfully persuade humans and amplify persuasive language. It is therefore essential to understand how user instructions affect the generation of persuasive language, and to understand whether the generated persuasive language differs, for example, when targeting different groups. In this work, we propose a framework for evaluating how persuasive language generation is affected by recipient gender, sender intent, or output language. We evaluate 13 LLMs and 16 languages using pairwise prompt instructions. We evaluate model responses on 19 categories of persuasive language using an LLM-as-judge setup grounded in social psychology and communication science. Our results reveal significant gender differences in the persuasive language generated across all models. These patterns reflect biases consistent with gender-stereotypical linguistic tendencies documented in social psychology and sociolinguistics.
Read the full abstract

All authors Amalie Brogaard Pauli, Maria Barrett, Max Müller-Eberstein, Isabelle Augenstein, Ira Assent

Computation and Language · Cross-listed from cs.AI

PII-VisBench: Evaluating Personally Identifiable Information Safety in Vision Language Models Along a Continuum of Visibility

G M Shahariar, Zabir Al Nazi, Md Olid Hasan Bhuiyan et al. (+1 more)

Vision Language Models (VLMs) are increasingly integrated into privacy-critical domains, yet existing evaluations of personally identifiable information (PII) leakage largely treat privacy as a static extraction task and ignore how a subject's online presence–the volume of their data available online–influences privacy alignment. We introduce PII-VisBench, a novel benchmark containing 4000 unique probes designed to evaluate VLM safety through the continuum of online presence. The benchmark stratifies 200 subjects into four visibility categories: high, medium, low, and zero–based on the extent and nature of their information available online. We evaluate 18 open-source VLMs (0.3B-32B) based on two key metrics: percentage of PII probing queries refused (Refusal Rate) and the fraction of non-refusal responses flagged for containing PII (Conditional PII Disclosure Rate). Across models, we observe a consistent pattern: refusals increase and PII disclosures decrease (9.10% high to 5.34% low) as subject visibility drops. We identify that models are more likely to disclose PII for high-visibility subjects, alongside substantial model-family heterogeneity and PII-type disparities. Finally, paraphrasing and jailbreak-style prompts expose attack and model-dependent failures, motivating visibility-aware safety evaluation and training interventions.
Read the full abstract

All authors G M Shahariar, Zabir Al Nazi, Md Olid Hasan Bhuiyan, Zhouxing Shi

Computation and Language

Visualising Information Flow in Word Embeddings with Diffusion Tensor Imaging

Thomas Fabian

Understanding how large language models (LLMs) represent natural language is a central challenge in natural language processing (NLP) research. Many existing methods extract word embeddings from an LLM, visualise the embedding space via point-plots, and compare the relative positions of certain words. However, this approach only considers single words and not whole natural language expressions, thus disregards the context in which a word is used. Here we present a novel tool for analysing and visualising information flow in natural language expressions by applying diffusion tensor imaging (DTI) to word embeddings. We find that DTI reveals how embedding space representations change between tokens. Tracking these changes within the layers of an LLM allows for comparing different model structures and could potentially reveal opportunities for pruning an LLM's under-utilised layers. Our results show that our visualisation method permits novel insights into how LLMs represent actual natural language expressions, extending the comparison of isolated word embeddings and improving the interpretability of NLP models.
Read the full abstract

Computation and Language

Multimodal In-context Learning for ASR of Low-resource Languages

Zhaolin Li and Jan Niehues

Automatic speech recognition (ASR) still covers only a small fraction of the world's languages, mainly due to supervised data scarcity. In-context learning (ICL) with large language models (LLMs) addresses this problem, but prior work largely focuses on high-resource languages covered during training and text-only settings. This paper investigates whether speech LLMs can learn unseen languages with multimodal ICL (MICL), and how this learning can be used to improve ASR. We conduct experiments with two speech LLMs, Phi-4 and Qwen3-Omni, on three diverse endangered languages. Firstly, we find that MICL is effective for unseen languages, leveraging both speech and text modalities. We further show that cross-lingual transfer learning improves MICL efficiency on target languages without training on them. Moreover, we analyze attention patterns to interpret MICL mechanisms, and we observe layer-dependent preferences between audio and text context, with an overall bias towards text. Finally, we show that prompt-based ASR with speech LLMs performs poorly on unseen languages, motivating a simple ASR system that combines a stronger acoustic model with a speech LLM via MICL-based selection of acoustic hypotheses. Results show that MICL consistently improves ASR performance, and that cross-lingual transfer learning matches or outperforms corpus-trained language models without using target-language data. Our code is publicly available.
Read the full abstract

Computation and Language · Cross-listed from cs.AI

Logic-Parametric Neuro-Symbolic NLI: Controlling Logical Formalisms for Verifiable LLM Reasoning

Ali Farjami, Luca Redondi and Marco Valentino

Large language models (LLMs) and theorem provers (TPs) can be effectively combined for verifiable natural language inference (NLI). However, existing approaches rely on a fixed logical formalism, a feature that limits robustness and adaptability. We propose a logic-parametric framework for neuro-symbolic NLI that treats the underlying logic not as a static background, but as a controllable component. Using the LogiKEy methodology, we embed a range of classical and non-classical formalisms into higher-order logic (HOL), enabling a systematic comparison of inference quality, explanation refinement, and proof behavior. We focus on normative reasoning, where the choice of logic has significant implications. In particular, we compare logic-external approaches, where normative requirements are encoded via axioms, with logic-internal approaches, where normative patterns emerge from the logic's built-in structure. Extensive experiments demonstrate that logic-internal strategies can consistently improve performance and produce more efficient hybrid proofs for NLI. In addition, we show that the effectiveness of a logic is domain-dependent, with first-order logic favouring commonsense reasoning, while deontic and modal logics excel in ethical domains. Our results highlight the value of making logic a first-class, parametric element in neuro-symbolic architectures for more robust, modular, and adaptable reasoning.
Read the full abstract

Computation and Language

Afri-MCQA: Multimodal Cultural Question Answering for African Languages

Atnafu Lambebo Tonja, Srija Anand, Emilio Villa-Cueva et al. (+16 more)

Africa is home to over one-third of the world's languages, yet remains underrepresented in AI research. We introduce Afri-MCQA, the first Multilingual Cultural Question-Answering benchmark covering 7.5k Q&A pairs across 15 African languages from 12 countries. The benchmark offers parallel English-African language Q&A pairs across text and speech modalities and was entirely created by native speakers. Benchmarking large language models (LLMs) on Afri-MCQA shows that open-weight models perform poorly across evaluated cultures, with near-zero accuracy on open-ended VQA when queried in native language or speech. To evaluate linguistic competence, we include control experiments meant to assess this specific aspect separate from cultural knowledge, and we observe significant performance gaps between native languages and English for both text and speech. These findings underscore the need for speech-first approaches, culturally grounded pretraining, and cross-lingual cultural transfer. To support more inclusive multimodal AI development in African languages, we release our Afri-MCQA under academic license or CC BY-NC 4.0 on HuggingFace (https://huggingface.co/datasets/Atnafu/Afri-MCQA)
Read the full abstract

All authors Atnafu Lambebo Tonja, Srija Anand, Emilio Villa-Cueva, Israel Abebe Azime, Jesujoba Oluwadara Alabi, Muhidin A. Mohamed, Debela Desalegn Yadeta, Negasi Haile Abadi, Abigail Oppong, Nnaemeka Casmir Obiefuna, Idris Abdulmumin, Naome A Etori, Eric Peter Wairagala, Kanda Patrick Tshinu, Imanigirimbabazi Emmanuel, Gabofetswe Malema, Alham Fikri Aji, David Ifeoluwa Adelani, Thamar Solorio

Computation and Language

Stephanie2: Thinking, Waiting, and Making Decisions Like Humans in Step-by-Step AI Social Chat

Hao Yang, Hongyuan Lu, Dingkang Yang et al. (+8 more)

Instant-messaging human social chat typically progresses through a sequence of short messages. Existing step-by-step AI chatting systems typically split a one-shot generation into multiple messages and send them sequentially, but they lack an active waiting mechanism and exhibit unnatural message pacing. In order to address these issues, we propose Stephanie2, a novel next-generation step-wise decision-making dialogue agent. With active waiting and message-pace adaptation, Stephanie2 explicitly decides at each step whether to send or wait, and models latency as the sum of thinking time and typing time to achieve more natural pacing. We further introduce a time-window-based dual-agent dialogue system to generate pseudo dialogue histories for human and automatic evaluations. Experiments show that Stephanie2 clearly outperforms Stephanie1 on metrics such as naturalness and engagement, and achieves a higher pass rate on human evaluation with the role identification Turing test.
Read the full abstract

All authors Hao Yang, Hongyuan Lu, Dingkang Yang, Wenliang Yang, Peng Sun, Xiaochuan Zhang, Jun Xiao, Kefan He, Wai Lam, Yang Liu, Xinhua Zeng

Computation and Language

Learning to Retrieve User History and Generate User Profiles for Personalized Persuasiveness Prediction

Sejun Park, Yoonah Park, Jongwon Lim et al. (+1 more)

Estimating the persuasiveness of messages is critical in various applications, from recommender systems to safety assessment of LLMs. While it is imperative to consider the target persuadee's characteristics, such as their values, experiences, and reasoning styles, there is currently no established systematic framework to optimize leveraging a persuadee's past activities (e.g., conversations) to the benefit of a persuasiveness prediction model. To address this problem, we propose a context-aware user profiling framework with two trainable components: a query generator that generates optimal queries to retrieve persuasion-relevant records from a user's history, and a profiler that summarizes these records into a profile to effectively inform the persuasiveness prediction model. Our evaluation on the ChangeMyView Reddit dataset shows consistent improvements over existing methods across multiple predictor models, raising F1 from 33% to 47% on Llama-3.3-70B-Instruct. Further analysis shows that effective user profiles are context-dependent and predictor-specific, rather than relying on static attributes or surface-level similarity. Together, these results highlight the importance of task-oriented, context-dependent user profiling for personalized persuasiveness prediction.
Read the full abstract

All authors Sejun Park, Yoonah Park, Jongwon Lim, Yohan Jo

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

Open World Knowledge Aided Single-Cell Foundation Model with Robust Cross-Modal Cell-Language Pre-training

Haoran Wang, Xuanyi Zhang, Shuangsang Fang et al. (+5 more)

Recent advancements in single-cell multi-omics, particularly RNA-seq, have provided profound insights into cellular heterogeneity and gene regulation. While pre-trained language model (PLM) paradigm based single-cell foundation models have shown promise, they remain constrained by insufficient integration of in-depth individual profiles and neglecting the influence of noise within multi-modal data. To address both issues, we propose an Open-world Language Knowledge-Aided Robust Single-Cell Foundation Model (OKR-CELL). It is built based on a cross-modal Cell-Language pre-training framework, which comprises two key innovations: (1) leveraging Large Language Models (LLMs) based workflow with retrieval-augmented generation (RAG) enriches cell textual descriptions using open-world knowledge; (2) devising a Cross-modal Robust Alignment (CRA) objective that incorporates sample reliability assessment, curriculum learning, and coupled momentum contrastive learning to strengthen the model's resistance to noisy data. After pretraining on 32M cell-text pairs, OKR-CELL obtains cutting-edge results across 6 evaluation tasks. Beyond standard benchmarks such as cell clustering, cell-type annotation, batch-effect correction, and few-shot annotation, the model also demonstrates superior performance in broader multi-modal applications, including zero-shot cell-type annotation and bidirectional cell-text retrieval.
Read the full abstract

All authors Haoran Wang, Xuanyi Zhang, Shuangsang Fang, Longke Ran, Ziqing Deng, Yong Zhang, Yuxiang Li, Shaoshuai Li