Computer Science -- Computation and Language source: export.arxiv.org/rss/cs.CL maintainer: @tmaehara.bsky.social https://bsky.app/profile/arxiv-cs-cl.bsky.socialhttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6hsd7whv2bJiaming Fan, Daming Cao, Canchen Huang, Jiale Fu, Jin Zhang, Junjie Gao, Kai Yang, Xiangzhong Luo, Xu Yang TreeGraft: Adaptive Multi-Drafter Grafting for Tree-Based Speculative Decoding https://arxiv.org/abs/2608.2611228 Aug 2026 23:01 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6hsd7whv2bhttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6hrfafg625Xinming Wang, Haoran Du, Yi Chen, Jian Xu, Hongming Yang, Han Hu, Yulong Chen, Cheng-Lin Liu, Xu-Yao Zhang ElementCheck: Complexity-Aware Long-Form Text Factuality Evaluation via Sentence Elements https://arxiv.org/abs/2608.2611828 Aug 2026 23:01 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6hrfafg625https://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6hqhf4zt2bArt Kanke DeflectBench: A Benchmark for Evaluating Rhetorical Fallacy Generation in LLMs https://arxiv.org/abs/2608.2611928 Aug 2026 23:00 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6hqhf4zt2bhttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6hpjpmcf2dJiajun He, Zongyu Guo, Jos\'e Miguel Hern\'andez-Lobato, Yuanqi Du Recipes for Steering and Scaling LLMs via Sampling https://arxiv.org/abs/2608.2612028 Aug 2026 23:00 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6hpjpmcf2dhttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6hk2lp4e2lAli Asaria, Tony Salomone, Deep Gandhi Can a Model Catch Its Own Hallucinations for Free?: Label-Free Doubt Signals Hold Their Own Against a Labelled Dataset for Abstention https://arxiv.org/abs/2608.2612128 Aug 2026 22:57 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6hk2lp4e2lhttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6hj4ftuh24Paarth Singh Rathore Which India Survives Translation? Narrative Homogenisation Across Indian Oral Traditions in LLMs https://arxiv.org/abs/2608.2612328 Aug 2026 22:56 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6hj4ftuh24https://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6hi6ox7o2jZiqiang Zhang, Jing Ma, Zilong Wang, Jiayuan Chen, Yi Qiao, Yu He, Wei Zhang, Dai Cheng, Xiaoyu Shen Natural-Language Policies to Executable Decisions: An Interpretable Large Language Model Framework https://arxiv.org/abs/2608.2612428 Aug 2026 22:56 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6hi6ox7o2jhttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6hhakb542sMuhammad Deedahwar Mazhar Qureshi, Sannaan Khan, Muhammad Atif Qureshi, Wael Rashwan Training-Time Explainability for Multilingual Hate Speech Detection: Aligning Model Reasoning with Human Rationales https://arxiv.org/abs/2608.2612528 Aug 2026 22:55 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6hhakb542shttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6hgcq2xi2gBohao Wang, Chenwei Wu, Haoyu Li, Hang Zou, Yu Tian, Lina Bariah, Li Wei, Chongwen Huang, Yongliang Shen, Zhaoyang Zhang, Merouane Debbah TelecomGPT-R1: A Unified Open-Source Reasoner for the Telecom Stack https://arxiv.org/abs/2608.2612628 Aug 2026 22:55 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6hgcq2xi2ghttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6hfex3sf2jPrabhjot Singh, Somnath Luitel, Manmeet Singh, Josh Durkee FIRSTPASS: A Multi-Domain, Multi-Round Peer Review Dataset Grounded in Real Editorial Outcomes https://arxiv.org/abs/2608.2612928 Aug 2026 22:54 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6hfex3sf2jhttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6hatipfg2eTatiana Petrova, Andrei Mazniak, Radu State Agents Don't Paginate: First-Chunk Selection for LLM Tool Responses https://arxiv.org/abs/2608.2613028 Aug 2026 22:52 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6hatipfg2ehttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6h7vt4mn25Ilija Subasic, Andrew Rabinovich, Zhao Chen Evaluating Language Models in Realistic Conversational Contexts https://arxiv.org/abs/2608.2613128 Aug 2026 22:51 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6h7vt4mn25https://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6h6xzkvm24Harish Karumuri, Mahesh Vemula, David Lopes Pegna Agent Seer: Synthesizing Scenarios from Specification Understanding https://arxiv.org/abs/2608.2613328 Aug 2026 22:51 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6h6xzkvm24https://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6h5zptn52bFrancesca von Braun-Bates, Sunreeta Sen, Indraayudh Talukdar, Anirban Lahiri Data Science Approaches to Evaluating Honours Candidates https://arxiv.org/abs/2608.2613528 Aug 2026 22:50 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6h5zptn52bhttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6h53t5gt23Tanvi Nagilla, Alexander Jameson, Daniel Manta, Shayaan Uddin Reward-Informed Sparse Autoencoders and the Solution-Completeness Confound https://arxiv.org/abs/2608.2613628 Aug 2026 22:50 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6h53t5gt23https://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6h45teih2eEichi Uehara Interpretable, Fairly Evaluated Automated L2 Speaking Assessment that Beats the Single-Human Ceiling and Why Pause Encoding Does Not Change LLM Fluency Scores https://arxiv.org/abs/2608.2613728 Aug 2026 22:49 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6h45teih2ehttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6gxbsh4523Rajveer Singh Pall, Sameer Yadav Cross-Platform Generalisation Failure in Mental Health Natural Language Processing: A Five-Axis Fairness Audit of Transformer Models on Social Media https://arxiv.org/abs/2608.2613828 Aug 2026 22:46 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6gxbsh4523https://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6gwd7ulm2bJiangrui Zhao, Xiaoting Du Syntax vs. Semantics: How Transformers Learn Deep Dependencies https://arxiv.org/abs/2608.2613928 Aug 2026 22:46 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6gwd7ulm2bhttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6gvdq4o625Kaihua Liang, An Zhong, Xin Tan, Zafar Ayyub Qazi, Hong Xu, Jian Weng, Marco Canini Affix Cache for Diffusion Large Language Models https://arxiv.org/abs/2608.2614028 Aug 2026 22:45 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6gvdq4o625https://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6gufyzxt2bZining Wang, Tongkun Guan, Boming Chen, Zhentao Guo, Jianqiang Liu, Chao Jin, Chen Duan, Kai Zhou, Pengfei Yan, Wei Shen, Xiaokang Yang AdaThinking-E: One-Token Entropy Regulation for Adaptive Thinking https://arxiv.org/abs/2608.2614128 Aug 2026 22:45 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6gufyzxt2bhttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6gtiiobc25Yuhan Liu, Yixiong Zou, Yuhua Li, Ruixuan Li Position Is All You Need: A Free Lunch Token Compression Strategy for MLLM-based Referring Expression Segmentation https://arxiv.org/abs/2608.2614228 Aug 2026 22:44 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6gtiiobc25https://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6gsl226i2bMuhammad Jawad Chowdhury, Adiba Hasan, Ishrak Hossain, Shahriar Ivan, Sabbir Ahmed Beyond Accuracy: A Qualitative Analysis of Vision-Language Models for Hate Speech Detection in Memes https://arxiv.org/abs/2608.2614328 Aug 2026 22:44 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6gsl226i2bhttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6go4tzpn2sSalima Lamsiyah, Ruslan Mitkov Why Current XAI Is Not Enough for Arabic NLP: A Critical Survey of the Explainability Gap https://arxiv.org/abs/2608.2614428 Aug 2026 22:41 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6go4tzpn2shttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6gn6c2rt2ePrathosh A P Vagdhenu: A Vrutta (Meter) Aware Shloka-to-Chant (TTS) System for Sanskrit https://arxiv.org/abs/2608.2614628 Aug 2026 22:41 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6gn6c2rt2ehttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6gmatz7n25Yucheng Zhou, Peng Luo, Qianning Wang, Chengzhong Xu, Jianbing Shen CARE: Causally-Aligned Reasoning Exploration for Medical Large Language Models https://arxiv.org/abs/2608.2614728 Aug 2026 22:40 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6gmatz7n25https://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6glde47b2sJonas L\"anzlinger, Katharina O. E. M\"uller, Burkhard Stiller, Bruno Rodrigues Towards Interpretable Depression Detection: Linking Acoustic Features to DSM-5 Indicators https://arxiv.org/abs/2608.2614828 Aug 2026 22:40 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6glde47b2shttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6gkfonc324Eric Lacosse, Mariana Duarte, Graham Todd, Peter M. Todd, Daniel C. McNamee Artificial Intelligence Models Can Predict and Collaboratively Modulate Human Memory Search https://arxiv.org/abs/2608.2615228 Aug 2026 22:39 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6gkfonc324https://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6gji4h2p2eMuhammad Aurangzeb Ahmad, Kim Shyu, Leon Oliver, Fergus Sleight, Paul Landau Evaluating AI Generated Summaries for Cancer Patients https://arxiv.org/abs/2608.2615428 Aug 2026 22:39 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6gji4h2p2ehttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6ge67qdi2bYixia Li, Yaqing Shi, Zhiwen Ruan, Dongdong Zhang, Lingjie Jiang, Shaohan Huang, Yun Chen, Guanhua Chen, Furu Wei VFA: Empowering Multilingual MLLMs via Vision-Free Adaptation https://arxiv.org/abs/2608.2615528 Aug 2026 22:36 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6ge67qdi2bhttps://bsky.app/profile/arxiv-cs-cl.bsky.social/post/3mu6gdaquk22gJesse St. Amand, Callum Canavan, Sohaib Imran, Joseph Hewson, Aaron Lutz, Shi Feng, Puria Radmard, Lennie Wells Self-Generated Text Recognition: Quality Heuristics, Cross-Task Transfer, and Downstream Bias in LLM Evaluation https://arxiv.org/abs/2608.2615928 Aug 2026 22:35 +0000at://did:plc:jf3oraummcsfodflx5w5pouf/app.bsky.feed.post/3mu6gdaquk22g