cs.CL · 2026-08-03 · No. 73

Computation and Language, 2026-08-03.

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

01 — The papers

15 entries
  1. 01

    TokTier: Exact Stateful Tokenization for Agentic LLM Serving

    Zhenyu Zhang, Zhichao Cao

    cs.CL · cs.DC · cs.PF

    LLM serving systems cache prompt KV state, yet most front ends still re-tokenize the full request text on every call. The cost lands on coding agents, which resubmit a long transcript after each small tool result, and reuse is hard because even a short append can change token boundaries near the end of the previous sequence. Across 153,951 calls from two agent ecosystems, the median call appends about 1.4K characters, and only 1.0-3.6% of...

    arxiv.org/abs/2607.29678 · PDF

  2. 02

    FriendBench: Benchmarking Dyadic Familiarity Inference in Humans and Multimodal Large Language Models

    Jeffrey M. Girard, Jason Z. Zheng, Jacqueline R. Vertino, Antony D'Avirro, Benjamin Peloquin

    cs.CL · cs.AI · cs.CV · cs.HC

    Reading a social situation often depends on behavior, not words alone. We introduce FriendBench, a benchmark for inferring whether two people are already familiar or are meeting as strangers, from a 20-second clip of a dyadic ice-breaker conversation. Every pair answers the same type of prompt, so only the manner of interaction can reveal the answer. Across text, audio, and video, we compare 26 models from seven companies against matched...

    arxiv.org/abs/2607.29602 · PDF

  3. 03

    ARB: A Matched Authorship-Rewriting Benchmark Dataset for AI-Text Detector Evaluation

    Gaetano Perrone, Simon Pietro Romano

    cs.CL · cs.AI

    Standard AI-text detection benchmarks compare human-written text against text generated directly by large language models (LLMs). While prior work has shown that rewriting and paraphrasing can degrade detector performance, it remains unclear whether performance measured on this conventional benchmark predicts detector behavior when human-authored content is rewritten by an LLM. To address this gap, we introduce Authorship-Rewriting Benchmark...

    arxiv.org/abs/2607.29539 · PDF

  4. 04

    Evidence-Type Competition: When Can Interventional Data Teach Language Models Causal Direction?

    Xining Xun

    cs.CL · cs.LG

    Interventional data is widely regarded as the gold standard for teaching models causal reasoning. We test this assumption in a fully controlled synthetic environment pitting observational correlation against causal effect, and find it fails instructively. In Simpson's-paradox worlds, where the two have systematically opposite signs, increasing the fraction of interventional samples in pretraining does not improve causal direction: the...

    arxiv.org/abs/2607.29484 · PDF

  5. 05

    PTP: Previous-Token Prediction based LLM Inversion for Near-Exact Prompt Reconstruction

    Pirzada Suhail, Nagasai Saketh Naidu, Atanu R Sinha, Amit Sethi

    cs.CL · cs.LG

    Large language models (LLMs) generate text by auto-regressively sampling the next token. This inherently leads to a many-to-many mapping between prompts and responses, complicating the task of inferring prompts from observed outputs. Prior work on LLM inversion frames prompt recovery as a semantic reconstruction task. They rely on fine-tuning pretrained sequence-to-sequence models on large external datasets--and requiring access to model...

    arxiv.org/abs/2607.29378 · PDF

  6. 06

    Cross-Lingual Transfer for Machine Translation in Turkic Languages

    Omer Burak Cinar, Mehmet Mert Dalkilic, Cagri Toraman

    cs.CL · cs.AI

    Cross-lingual transfer is central to low-resource machine translation, but its behavior within closely related language families remains insufficiently characterized. We study transfer among five Turkic languages; Turkish, Azerbaijani, Uzbek, Kazakh, and Kyrgyz; using pairwise transfer matrices. In this setting, each model is fine-tuned with one transfer source and evaluated on a different transfer target while the translation target remains...

    arxiv.org/abs/2607.29355 · PDF

  7. 07

    Translation with Thought: Difficulty-Adaptive Reasoning via Reinforcement Learning for Multi-Domain Machine Translation

    Yongshi Ye, Biao Fu, Chongxuan Huang, Yidong Chen, Xiaodong Shi

    cs.CL · cs.AI

    Multi-domain machine translation (MDMT) poses a unique challenge due to varying levels of linguistic complexity across domains. Inspired by human translators' ability to adapt reasoning effort based on difficulty, we propose TwT (Translation with Thought), a resource-rational framework that learns to modulate inference between intuitive and deliberate reasoning. TwT is trained in two stages: (1) supervised fine-tuning on difficulty-aware long...

    arxiv.org/abs/2607.29287 · PDF

  8. 08

    CalibratedRubric: Task-Adaptive Rubric Banks for Open-Ended LLM Evaluation

    Mengting Chen, Yanshu Sun, Wanting Liang, Beidi Luan, Rui Sun, Dezhi Chen, Jing Li, Zuo Bai

    cs.CL · cs.AI · cs.LG

    Reliable evaluation of open-ended LLM outputs requires fine-grained rubrics, yet expert curation is costly and difficult to scale. Existing automated pipelines rely on strict judge unanimity and binary variance filters, which cannot distinguish measurable rubrics from informative ones. We introduce CalibratedRubric, a task-adaptive framework that combines type-specific scoring, Bayesian rubric-measurability filtering, and item response theory...

    arxiv.org/abs/2607.29252 · PDF

  9. 09

    Small Is Enough: Per-User Style Rewriting of AI-Edited Text via LoRA Adapters

    Antorweep Chakravorty

    cs.CL · cs.AI · cs.CY · cs.ET · cs.HC

    InMyStyle is a privacy first, single user system that adapts small language models to rewrite AI-edited text towards an individual user's writing style without an instruction prompt at inference. Given a user's documents, it uses multiple local helper LLMs to construct paired training examples and fine tunes LoRA adapters on base models ranging from 0.5B to 7B parameters. Length aware generation budgets and automatic chunking support inputs...

    arxiv.org/abs/2607.29238 · PDF

  10. 10

    Semantics of Subterfuge: Benchmarking Legal Deception Detection Against General-domain State-of-the-Art

    Theekshana Samaradiwakara, Nisansa de Silva, George C. Lobb

    cs.CL · cs.AI

    Deception detection has critical implications for legal proceedings, law enforcement, and online security. Although human judgment is limited in accuracy and scalability, Natural Language Processing (NLP) offers a data-driven alternative. We present a survey and comparative analysis of NLP-based Automatic Deception Detection (ADD) focusing on the legal domain, reviewing the evolution from feature-based machine learning to Large Language Model...

    arxiv.org/abs/2607.29066 · PDF

  11. 11

    PARALLEL: A Prefrontal-Aligned Reinforcement inspired Approach for Language-Model Learning under Explicit Limits

    Namkyung Yoon, Sanghong Kim, Hwangnam Kim

    cs.CL · cs.AI

    Recent language models achieve strong performance across a variety of tasks, but conventional adaptation applies updates uniformly across training samples regardless of their local update benefit. We propose PARALLEL, a prefrontal-aligned reinforcement inspired approach for language-model learning. Inspired by the complementary roles of goal-related and uncertainty-related control, PARALLEL represents these forms of information as separate...

    arxiv.org/abs/2607.28982 · PDF

  12. 12

    FairFund-Bench: Evaluating Distributive Bias in LLM Resource Allocation

    Martin Lukk

    cs.CL · cs.AI · cs.CY

    Large language models (LLMs) are increasingly involved in the distribution of scarce resources, raising concerns about biased allocations based on characteristics like race and gender. Recent LLM audits have produced inconsistent results, however, finding evidence of both positive and negative discrimination towards women and ethnic minorities, even for the same models. We show that this disagreement can arise from differences in audit format...

    arxiv.org/abs/2607.28934 · PDF

  13. 13

    TextCloak: Thwarting Unauthorized LLM Exploitation via RL-Driven Unlearnable Text

    Chengshuai Zhao, Pingchuan Ma, Dawei Li, Bohan Jiang, Zhiyuan Yu, Zhen Tan, Huan Liu

    cs.CL · cs.AI · cs.CR · cs.LG

    The rapid development of Large Language Models (LLMs) has led to significant advances across a wide range of language tasks, while simultaneously raising growing concerns about unauthorized data exploitation and privacy leakage. Unlearnable examples (UEs) offer a promising defense by introducing carefully designed perturbations into data such that models trained on them exhibit degraded utility. However, existing methods for text protection...

    arxiv.org/abs/2607.28862 · PDF

  14. 14

    Rolling With Resistance: Preference-Optimized LLM Counselors Can Trade Goal Persistence for Relational Attunement in Motivational Interviewing

    Weiying Chen, Junlong Shen, Zhexuan Tang

    cs.CL · cs.AI · cs.LG

    In Motivational Interviewing (MI), a client's sustain talk (arguments for the status quo) calls for the counselor to roll with resistance, a move that can fail in two opposite ways: capitulation (abandoning the change agenda to preserve rapport) or confrontation (arguing or directing, overriding the client's autonomy). We introduce a two-axis evaluation of counselor responses, anchored in the Motivational Interviewing Treatment Integrity...

    arxiv.org/abs/2607.28814 · PDF

  15. 15

    Benchmarks Are Not Monolithic: Sample-Level Auditing and Orchestration for LLM Evaluation

    Philipp D. Siedler, Jordan Sassoon

    cs.CL · cs.AI · cs.LG

    Benchmark datasets are central to evaluating Large Language Models (LLMs), yet they are typically conceived as monolithic tasks, obscuring substantial variation in the demands of individual samples. We introduce a dataset-centric meta-evaluation framework that audits benchmark datasets at the sample level along five latent dimensions: 1. Cognitive and Knowledge Demands, 2. Language and Content Quality, 3. Task Properties, 4. Context, and 5....

    arxiv.org/abs/2607.28801 · 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.