cs.CL · 2026-11-02 · No. 161

Computation and Language, 2026-11-02.

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

01 — The papers

13 entries
  1. 01

    Subquadratic Attention with a Hierarchical Tree-Structured Indexer

    Leon Lufkin, Cutter Dawes, Alexander Speicher, Kamesh Krishnamurthy, Tomás Figliolia, Beren Millidge

    cs.CL · cs.AI

    The quadratic scaling of attention is a fundamental bottleneck limiting the long-context capabilities of large language models (LLMs). A variety of approaches have attempted to circumvent this, but they either fail to match attention's expressiveness (e.g., sliding-window or linear attention variants with fixed-size state) or only achieve constant-factor speedups (e.g., sparse attention methods whose indexers themselves scale quadratically)....

    arxiv.org/abs/2610.42112 · PDF

  2. 02

    CRoS: Causality Reward on Scrambled CoTs Improves Accuracy and Reasoning Faithfulness

    Xiao Yu, Justin Chih-Yao Chen, Zaid Khan, Hyunji Lee, Mohit Bansal

    cs.CL · cs.AI

    Reinforcement learning with verifiable rewards (RLVR) has become the standard method for training language models to reason. However, RLVR typically relies solely on outcome correctness, which does not guarantee that the CoT causally contributes to the final answer. Without this connection, outcome rewards may fail to reinforce high-quality reasoning. Existing causality metrics assess CoT causality by perturbing the CoT, such as deleting...

    arxiv.org/abs/2610.42054 · PDF

  3. 03

    Aligning MLLMs with the Latent Structure of Human Cognition via Behavior-Derived Semantic Dimensions

    Ning E, Changde Du, Yizhuo Lu, Huiguang He

    cs.CL · cs.AI

    Multimodal Large Language Models (MLLMs) have achieved impressive performance across a wide range of reasoning tasks, yet it remains unclear whether their internal representations and decision criteria align with the latent structure of human cognition. In this work, we propose a novel framework to align MLLMs with the multidimensional mental representations underlying human similarity judgements. We build on a previously established...

    arxiv.org/abs/2610.41951 · PDF

  4. 04

    Text as an Index for Zero-Shot Cross-Domain Knowledge Tracing

    Yin Yi, Haiping Ma, Siyu Song, Ziwen Wang, Shangshang Yang

    cs.CL · cs.CY · cs.LG

    Knowledge tracing (KT) predicts a student's next response from their history. New platforms often have question texts but no response logs for training, motivating zero-shot cross-domain KT. Existing transfer and cold-start methods use question semantics, but cross-platform semantic discrepancies limit transferability. We propose TIZSKT, which separates behavioral sequence modeling from domain-local historical evidence construction and uses...

    arxiv.org/abs/2610.41883 · PDF

  5. 05

    SimXAI: A Simple Drop-in Layer Increases Interaction Quality and Understanding in Conversational XAI

    Van Bach Nguyen, Qianli Wang, Nils Feldhus, Jörg Schlötterer, Christin Seifert

    cs.CL · cs.AI · cs.HC

    Conversational explainable AI (ConvXAI) agents often use GUS-style architectures with separate modules for intent recognition, dialogue management, and template-based response generation. While faithful to the underlying XAI methods, these pipelines are hard to adapt and often struggle with context-dependent follow-up questions. We present SimXAI, a lightweight LLM-centric conversational layer that replaces these modules with a single...

    arxiv.org/abs/2610.41856 · PDF

  6. 06

    Peeling the Matryoshka: Extracting Ontological Structure from Retrieval Embeddings

    Eunchan Kim, Inwhee Joe

    cs.CL · cs.IR · cs.LG

    Taxonomy construction is essential for knowledge-intensive NLP yet remains costly: clustering-based methods (i) need careful per-corpus tuning and yield unstable hierarchies, while LLM-based pipelines (ii) depend on expensive task-specific training that does not transfer across domains. We report an empirical finding that bears on both: in Matryoshka Representation Learning (MRL) embeddings, where each leading prefix is itself a usable,...

    arxiv.org/abs/2610.41785 · PDF

  7. 07

    When Steering Shifts Discourse: Identifying and Intervening on Discourse-Level Changes due to Activation Steering in LLMs

    Shivam Adarsh, Hetianyu Huang, Lau Safin, Maria Maistro, Christina Lioma

    cs.CL · cs.AI

    Activation steering is an inference-time technique used to control or modify the behavior, tone, or safety properties of a large language model (LLM) without retraining the model's weights or editing its input prompt, but simply by adding a vector to its residual-stream activations. However, activation steering also changes the generated text in unintended ways, unrelated to intended behavior change, that are not well understood. Prior work...

    arxiv.org/abs/2610.41661 · PDF

  8. 08

    BYOD: Build Your Own DLM - Efficient Autoregressive-to-Diffusion Language Model Conversion

    Ruurd J. A. Kuiper, Robert A. Bagheri

    cs.CL · cs.LG

    Unlike autoregressive (AR) models, diffusion language models (DLMs) permit parallel token updates, flexible generation, and bidirectional attention, but leading systems rely on costly pretraining or full-model fine-tuning. We present Build Your Own DLM (BYOD), a generalizable LoRA-only recipe for parameter-efficient AR-to-DLM conversion. Without architecture-specific tuning, BYOD induces diffusion generation in Gemma 2 9B, Llama 3.1 8B,...

    arxiv.org/abs/2610.41574 · PDF

  9. 09

    Sparse Text Diffusion under Dual Budgets: Context Connectivity and Expert Activation

    Alexandros Christoforos, Chadbourne Davis, Paul Raj

    cs.CL · cs.AI

    Long-context diffusion repeatedly pays for both token interaction and feed-forward computation. We formulate these costs as two separate budgets: the edges available to an attention graph and the experts activated for each token. This view connects local, dilated, and global attention with conditional expert computation without assuming that sparsity alone preserves document-level information. We specify normalized masked attention,...

    arxiv.org/abs/2610.41487 · PDF

  10. 10

    KEEPER: Learning Control-Sufficient Textual Memory for Long-Horizon LLM Agents

    Jinkyeong Choi, Chaebin Jeong, Junsung Jeon, Donghyeon Park

    cs.CL · cs.AI

    Recursive textual memory can carry earlier errors into later updates, yet task performance alone does not establish memory reliability. We define self-consistency bias as the persistence of incorrect claims already present in the incoming memory. We present KEEPER, a comparative analysis of four auxiliary objectives along information-content and update-sensitivity axes within a shared recursive interface, examining how supervision targets...

    arxiv.org/abs/2610.41482 · PDF

  11. 11

    KV-Cache Patching: Tracing How and Where Fine-Tuned Behaviors Arise

    Jonas Loos, Oliver Eberle

    cs.CL · cs.LG

    Fine-tuning a language model changes how it responds, but where is this fine-tuned behavior instantiated? To better understand this, we evaluate whether the change in model behavior arises only during the generation of the answer tokens where the change manifests, or whether altered processing of previous tokens causes it via the attention mechanism. For this, we patch the KV cache from a LoRA fine-tuned model into the unmodified base model,...

    arxiv.org/abs/2610.41480 · PDF

  12. 12

    BlindToChange: AI Agents Can Find Changes When Asked But Do They Look When Needed?

    Anirudh Phukan, Tanya Goyal, Danish Pruthi

    cs.CL · cs.AI

    Given that AI agents can search the web broadly and deeply in real time, it is tempting to conclude that their outputs account for recent world events. We show that this conclusion does not hold in practice. While agents can search and easily find recent changes, the correctness of their responses depends on whether they search and what they search for. Popular agentic search benchmarks like BrowseComp cannot expose failures wherein agents do...

    arxiv.org/abs/2610.41435 · PDF

  13. 13

    In-Model Context Compression with Agentic Soft Tokens for Long-Horizon LLM Agents

    Xuning He, Kai Fan, Runyu Zhang, Binghuan Wang, Yuxiang Zhang, Xue Jiang, Yihong Dong

    cs.CL · cs.AI · cs.LG · cs.NE

    Long-horizon LLM agents play an important role in solving complex and sustained tasks, with ever-growing interaction contexts. However, these accumulated contexts will raise substantial computational cost and obscure key decision-relevant evidence. Existing methods typically train external compressors or context-management agents to rewrite context as shorter hard-token sequences, but such external selection may not align with the LLM's...

    arxiv.org/abs/2610.41291 · 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.