cs.CL · 2026-10-20 · No. 149

Computation and Language, 2026-10-20.

16 new papers in cs.CL. Titles, authors, abstracts. Links to arXiv. Want this in your inbox every morning? Subscribe →

01 — The papers

16 entries
  1. 01

    OpenVLHarness: Real World Visual Agentic Tool-Use with Multimodal Memory

    Yu Zhou, Cheng-Fu Yang, Rui Sun, Di Wu, Sibo Peng, Mingyang Zhang, Zhen Yang, Ruiyi Zhang, Nanyun Peng, Zhe Gan,...

    cs.CL · cs.AI · cs.CV · cs.LG · cs.MM

    Multimodal large language models (MLLMs) have made rapid progress on visual reasoning, yet many tasks require evidence that is difficult to obtain through internal reasoning alone: our controlled study on 22 benchmarks shows that simply increasing the inference budget of the base model yields only modest gains. Such tasks demand grounded perception, external knowledge, and numerical measurements obtained through heterogeneous tools, whose...

    arxiv.org/abs/2610.26860 · PDF

  2. 02

    LatComm: Reusable Context Compression for Latent Communication between Large Language Models

    Huawei Lin, Jinting Liu, Tan Yu, Weijie Zhao

    cs.CL · cs.AI

    Systems that combine several large language models (LLMs) often use one context for many questions. When the context is passed between models as text, the receiving model processes all of its tokens again for every question. Avoiding this cost requires a context representation that is compact, reusable across questions, and readable by a different LLM. Existing compression and latent communication methods each provide only some of these...

    arxiv.org/abs/2610.26840 · PDF

  3. 03

    RefCraft: Benchmarking Visual-Reference-Conditioned Agentic Generation through Web Development

    Yuchi Wang, Rongyao Fang, Weihao Yu, Xilin Wei, Tiancheng Gu, Dingkang Yang, Hang Zhang, Dayiheng Liu, Hongsheng Li,...

    cs.CL · cs.AI · cs.CV

    Beyond basic tool use and coding skills, effectively incorporating visual references into agentic generation is equally essential for applying AI agents to real-world productivity tasks. We study this crucial capability through webpage development, a practical and readily evaluable setting in which existing benchmarks remain largely idealized: they typically focus on reconstructing visual targets or generating webpages from textual...

    arxiv.org/abs/2610.26826 · PDF

  4. 04

    MemT-1.0: Towards Memory-Native Evolution

    Zhiyu Li, Zeyu Zhang, Linyun Liu, Yiwei Bu, Chenyang Xi, Xichong Zhang, Yi Yu, Yu Yu, Zhichen Liu, Yuxiao Yi, Chunyu...

    cs.CL · cs.LG

    Memory-native evolution requires making memory an intrinsic capability of foundation models, enabling them to continuously accumulate and utilize information. To this end, we introduce MemT-1.0, a new architecture of memory foundation models that internalize memory as part of model computation. Building on Metis, MemT-1.0 employs Gated DeltaNet-2 with a stable token-level memory update mechanism within each turn. We further introduce...

    arxiv.org/abs/2610.26825 · PDF

  5. 05

    AfriBED: a Speech Recognition Bias Evaluation Dataset for Low Resource African Languages

    Claytone Sikasote, Hussein Suleman, Jan Buys

    cs.CL · cs.AI

    Despite advancements in automatic speech recognition systems (ASR), current systems exhibit speaker-attribute bias, i.e., their speech recognition performance is biased against certain speakers based on their demographic attributes such as gender, accents, and speaker type. However, accurately assessing speaker attribute bias in a given ASR system requires the availability of a bias evaluation dataset with comparable speech across subgroups...

    arxiv.org/abs/2610.26821 · PDF

  6. 06

    TokenGate: Adapting LLMs for Fast and Memory-Efficient Prefill with Learned Token Pruning

    Liana Patel, Rishabh Ranjan, Ion Stoica, Matei Zaharia, Carlos Guestrin

    cs.CL · cs.AI

    Expert human readers process large inputs (e.g., code repositories and long documents) by skimming to closely read only what matters. On the other hand, language models process all input tokens through each model layer, resulting in significant processing costs for long-context inputs. We introduce TokenGate, a small, drop-in component that extends existing trained models to enable end-to-end learnable token pruning over long-context inputs...

    arxiv.org/abs/2610.26754 · PDF

  7. 07

    Hierarchical Token Merging with Recursive Autoencoders for Unsupervised Grammar Induction

    Murat Guney Kemal, Robert Lieck

    cs.CL · cs.AI

    Hierarchical abstraction is a fundamental principle across many data modalities, enabling complex sequences to be represented compositionally at multiple levels. Learning such hierarchies directly from data, without explicit structural supervision, remains a central challenge in representation learning. We investigate whether meaningful hierarchical structure can emerge from compression alone. While autoencoders learn compressed...

    arxiv.org/abs/2610.26695 · PDF

  8. 08

    The Development of Sycophancy in Language Models: Tracing Vulnerabilities from Pretraining to Post-Training

    Haonan Huang, Chengkun Cai

    cs.CL · cs.AI · cs.LG

    Language models can change a committed answer to accommodate a user's opinion or pushback without new evidence. However, when this behaviour develops during training and what a released model inherits from earlier stages remain unclear. We follow 741 checkpoints of 30 open model lineages, from early pretraining to released models. We plant initial verdicts on moral scenarios and measure how far a model follows a user's stated opinion or...

    arxiv.org/abs/2610.26655 · PDF

  9. 09

    Conducting Reliable Keyword Estimation with Unreliable ASR in Police Audio Data

    Robert Cole Molloy, Rhiannon N. Miller, Stephen Morgan, Anjalie Field

    cs.CL · cs.AI

    Extracting insights from large audio data sets using human transcription is expensive and time-consuming. Off-the-shelf Automatic Speech Recognition (ASR) systems offer an easy-to-use scalable alternative, but systematic transcription errors, especially in domain-specific settings, can bias downstream analyses and lead to unreliable conclusions. We investigate whether imperfect ASR transcripts can still be used to obtain statistically valid...

    arxiv.org/abs/2610.26614 · PDF

  10. 10

    Mitigating Reward Hacking in Preference Reward Models with Evolving Penalties

    Afonso Baldo, António Farinhas, Marcos Treviso, Ricardo Rei, Nuno M. Guerreiro

    cs.CL · cs.AI

    Reinforcement learning for language models has moved from preference reward models to verifiable rewards, which sustain much longer optimization. Preference models are prone to reward hacking: the policy raises their score without improving response quality, or even while degrading it. Yet for open-ended tasks, where quality can neither be verified nor fully written down as rubrics, they remain valuable. We introduce Dynalize, which mitigates...

    arxiv.org/abs/2610.26496 · PDF

  11. 11

    Co-Evolving Policy and Reward for Code Generation: A Mechanistic Study of Single-Adapter Drift

    Amarsaikhan Batjargal, Noriko Tomuro

    cs.CL · cs.LG · cs.SE

    Reinforcement learning with verifiable rewards (RLVR) is particularly well suited to code generation because generated programs can be objectively evaluated through execution. Execution also provides behavioral information that can be reused as additional supervision. The Synergistic Policy and Reward Co-Evolving Framework (SPARK), originally proposed for reasoning tasks, exploits this idea by recycling training rollouts and their evaluation...

    arxiv.org/abs/2610.26416 · PDF

  12. 12

    Large-Scale Biomedical RAG: Corpus Scaling, Document Representation, and Efficiency Trade-offs

    Adrián Sánchez-Freire, Gonzalo Pérez-Gómez, Aitziber Atutxa, Ander Barrena

    cs.CL · cs.AI · cs.IR

    We present a controlled, large-scale evaluation of how corpus scale, document representation, and reranking affect the effectiveness and efficiency of biomedical Retrieval-Augmented Generation (RAG) systems. Retrieval collections were progressively expanded while retaining the relevant source documents, enabling consistent comparisons across corpus sizes. We compared abstract, flat full-text, and hierarchical full-text representations across...

    arxiv.org/abs/2610.26339 · PDF

  13. 13

    VEILED-SH: Vulnerability Expression through Indirect Language Evaluation Dataset (Self-Harm)

    Leanne Tan, Keith Chua, Shaun Khoo, Onno P. Kampman, Lim Ruling Charmaine, Hazirah Hoosainsah, Kellie Yu Hui Sim,...

    cs.CL · cs.AI · cs.CY

    AI systems may miss indirect, culturally situated distress or provide support mismatched to the application context. Existing benchmarks do not jointly address cultural variation, application context, and practitioner disagreement. We introduce VEILED-SH, a benchmark and construction methodology for evaluating single-turn responses to indirect expressions of suicide, self-harm, and distress. Its 10,003 prompts span counselor chat, forum...

    arxiv.org/abs/2610.26190 · PDF

  14. 14

    Honeyed Words, Hidden Daggers: Benchmarking and Training LLMs to See Through Mock Politeness

    Caixin Kang, Sitong Gong, Tianyu Yan, Kaipeng Zhang, Yoichi Sato, Yifei Huang

    cs.CL · cs.AI

    The Chinese idiom "kou mi fu jian" ("honeyed words, hidden daggers") names speech that is polite in form but hostile in function. We present ZhenhuanBench, a 1,592-item benchmark built from expert commentary on Empresses in the Palace, covering speaker intent, emotion-language mismatch, and per-listener uptake. Text-only evaluation covers 34 LLMs. GPT-6 Astra leads the leaderboard at 73.48% overall, while its listener-reaction score is...

    arxiv.org/abs/2610.26059 · PDF

  15. 15

    Accelerated Drift of Scientific Prose in the LLM Era: Evidence from an Interrupted Time-Series Analysis

    Hanna Hubarava, Benjamin V. Ineichen

    cs.CL · cs.AI

    Every scientific text is embedded in the writing norms of its time: it reflects earlier publications while also shaping the texts that follow. The growing use of LLM-based writing tools introduces a new factor in this evolving system, but the linguistic changes coinciding with their emergence remain insufficiently understood. We investigate whether the trajectories of scientific-preprint writing changed in rate or direction after...

    arxiv.org/abs/2610.26008 · PDF

  16. 16

    TempoMoE: Identifying and Steering Fast-Slow Thinking Experts for Efficient Multimodal Reasoning

    Xiyue Zhang, Keru Zhou, Ziyi Wang, Bing Hu, Li Wang

    cs.CL · cs.CV · cs.LG

    Multimodal Mixture-of-Experts (MoE) reasoning models rely on long chain-of-thought (CoT) for complex tasks, yet often overthink easy inputs, increasing token cost and latency, and sometimes degrading accuracy. Existing methods optimize reasoning length during training or impose external control at inference, but leave the underlying reasoning mechanism a black box. We find that fast and slow thinking exhibit distinct expert-routing patterns:...

    arxiv.org/abs/2610.25999 · PDF

This edition is part of The Daily Abstract — cs.CL archive. Subscribe to receive these in your inbox each morning, automatically translated to Spanish, with reply-to-PDF: arxivdaily.ignorelist.com.

Colophon Set in Georgia, with system sans for interface chrome and a monospaced stack for code and paper identifiers. Sole accent: amber #D99C5E. Built and served on an always-free VM. The masthead is set 14% letterspaced because newspapers do that and it works.