cs.CL（2026-08-31）

📊 共 53 篇论文 | 🔗 7 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (43 🔗6) 支柱二：RL算法与架构 (RL & Architecture) (7 🔗1) 支柱八：物理动画 (Physics-based Animation) (2) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (43 篇)

#题目一句话要点标签🔗⭐
1 Where Identity Lives: Localized, Retain-Free Identity Unlearning in Multimodal Large Language Models 提出PAVA方法以解决多模态大语言模型中的身份信息删除问题 large language model multimodal
2 MMDS-Bench: Benchmarking Multimodal Large Language Models on Dynamic Stance in Social Media Interactions 提出MMDS-Bench以解决社交媒体动态立场分类问题 large language model multimodal
3 Reading the News: Adapting Large Language Models to Swedish Journalism Through Continued Pre-Training 通过继续预训练适应大型语言模型于瑞典新闻领域 large language model instruction following
4 LCoT-GV: Graph Attention Networks for Verifying Long Reasoning Chains in Large Language Models 提出LCoT-GV以验证大型语言模型的长推理链 large language model chain-of-thought
5 When LLM Meets Tree Search: A Systematic View of Inference as Search in Large Language Models 提出树搜索方法以优化大语言模型推理过程 large language model chain-of-thought
6 Quantifying and Mitigating Korean Jamo-Level Typographical Vulnerabilities in Large Language Models 提出Typo-Aware Chain-of-Thought以解决韩文Jamo级别的输入错误问题 large language model chain-of-thought
7 ImageEval 2026: Culturally Grounded Arabic Multimodal Evaluation 提出ImageEval 2026以解决阿拉伯多模态评估问题 multimodal
8 Every Token Leaves a Ripple in the Stream of Thought: Eliciting Model-Internal Token Saliency for Chain-of-Thought Compression 提出MIST以解决链式推理压缩中的令牌选择问题 chain-of-thought
9 SwarmBench: Can Large Language Models Act as Agent Swarm Orchestrators? 提出SwarmBench以评估大语言模型在多智能体编排中的能力 large language model
10 Sequential Trajectories and Simultaneous Blending: Multi-Emotion Modeling for Instruction-Following TTS 提出HybridEmo以解决多情感控制的TTS问题 instruction following
11 Read the Room, Read the Image: Understanding Indirect Speech Acts in Multimodal Visual Contexts 提出READI基准以解决间接言语行为理解问题 multimodal
12 GPAgentBench-2K: Benchmarking Large Language Model Agents in Complex Clinical Action Space 提出GPAgentBench-2K以解决临床决策中的行动空间限制问题 large language model
13 BiG-SURE - Bipartite Graph for Semantic Uncertainty and Reliability Estimation of LLMs 提出BiG-SURE以解决LLMs的不确定性估计问题 large language model multimodal
14 Evidence-Bounded Mental Health Reasoning from Heterogeneous Speech Protocols 提出基于证据的心理健康推理框架以解决多模态数据的有效性问题 multimodal chain-of-thought
15 GUIDE: Guiding Internal Evidence with Language Instructions 提出GUIDE框架以调控多模态模型的内部证据使用 multimodal instruction following
16 OCR-MetaReasoning Benchmark: Evaluating the Meta-Reasoning Ability of MLLMs in Text-Rich Image Understanding 提出OCR-MetaReasoning基准以评估MLLMs在文本丰富图像理解中的推理能力 large language model multimodal ✅
17 Reactivating Test-Time Scaling for Plane Geometry Problem Solving 提出多轨合成方法以解决平面几何问题的推理挑战 multimodal visual grounding ✅
18 SocialReasonBench: A Video-QA Benchmark for Social Reasoning with Counterfactual Narrative Videos 提出SocialReasonBench以解决视频社交推理评估问题 multimodal
19 Lies We Can See: Joint Verbal and Non-Verbal Deception by VLM Agents in Embodied Social Interactions 提出MineAmongUs以解决多模态社交互动中的欺骗问题 multimodal
20 TopoCompress: Long Context Compression via Graph-Wired Semantic Trajectories 提出TopoCompress以解决长上下文压缩问题 large language model
21 An Agentic Retrobiosynthesis Framework with Learned Frontier Selection 提出基于学习的边界选择框架以优化反向生物合成 large language model
22 Two Centuries of Sexism in British Parliament: A Computational Analysis of Women's Representation in the Hansard Corpus 通过计算分析揭示英国议会两百年的性别歧视现象 large language model
23 Beyond Token-Level Guidance: Inference-Time Alignment of Specialized LLMs via Cross-Family Representation Steering 提出CREST以解决专用LLM推理时安全性问题 large language model ✅
24 Using Prosody to Predict Syntactic Structure 提出信息论框架以量化韵律与句法结构的关系 multimodal
25 Calibrating Small Language Models for Claim Check-Worthiness Detection 提出NN-PPI以解决小型语言模型的准确性问题 large language model
26 Hidden Threat in Synthetic Data: Covert Targeted Bias Injection through Benign Text 提出隐蔽偏见注入方法以解决合成数据安全问题 large language model
27 Enhancing Low-Resource Language Reasoning via High-Resource Language Feature Transfer 通过高资源语言特征转移提升低资源语言推理能力 large language model
28 Beyond Surface Forms: Symbolic Edits as a Test for Logical Reasoning with LLMs 提出符号编辑框架以评估LLMs的逻辑推理能力 large language model
29 The Differential Reasoning Router: Operationalizing Cost-Aware LLM Annotation in E-commerce 提出差异推理路由器以解决电商冷启动标注问题 large language model
30 CPR for LLMs: Critical-Point Routing against Catastrophic Forgetting in Domain Adaptation 提出CPR以解决大语言模型领域适应中的灾难性遗忘问题 large language model
31 Low-Resource Preference Adaptation of LLMs via Activation-Based Label Propagation 提出激活基于标签传播的方法以解决低资源偏好适应问题 large language model ✅
32 You Shouldn't Have Asked: A Pragmatics-Inspired Taxonomy for Evaluating LLM Refusals 提出基于语用学的分类法以评估大型语言模型的拒绝行为 large language model
33 Error-Type-Aware Loss Reweighting for Robust Named Entity Recognition with Noisy LLM Labels 提出错误类型感知损失重加权以解决NER中的噪声标签问题 large language model
34 CLIN: an Objective Framework for Evaluating Creativity in Short Persian Literary Text 提出CLIN框架以评估波斯短文学文本的创造力 large language model
35 MURANO: Design, Run, and Reproduce Mechanistic Interpretability Experiments as Composable Pipelines 提出Murano框架以解决机制可解释性研究的整合问题 large language model
36 More Capable, Less Faithful: A Multilingual Analysis of Mathematical (Un)Solvability Detection in LLMs 提出多语言基准以解决大语言模型数学可解性检测问题 large language model
37 From Final Artifacts to Trajectories: Retrospective Process Supervision for Evidence-Grounded Long-Form Generation 提出RetroGen框架以解决开放任务中的轨迹数据不足问题 large language model
38 Learning to Reason and Use Tools through Unsupervised Fine-Tuning in Task-Oriented Dialog Systems 通过无监督微调提升任务导向对话系统的推理与工具使用能力 large language model
39 AIA$^{2}$: Attribute-Agnostic Imbalance Augmentation for Subgroup Robustness 提出AIA²框架以解决子群体鲁棒性问题 large language model ✅
40 When Errors Become Memories: Causal Pathway Tracing in Multi-Turn Memory-Augmented LLMs 提出基于结构因果模型的框架以解决记忆增强LLM中的错误传播问题 large language model
41 ALTSTEER: Selective Safety Steering for Moving Beyond Hard Refusals to Constructive Alternatives 提出ALTSTEER以解决安全引导中的拒绝问题 large language model
42 CAST: Critique-Aware Supervision for Training Reliable Long-Horizon Tool-Calling Agents 提出CAST框架以解决长时间工具调用代理的可靠性问题 large language model
43 Verification-Aware Training for Speculative Decoding 提出验证感知训练以提升推测解码效率 large language model ✅

🔬 支柱二：RL算法与架构 (RL & Architecture) (7 篇)

#题目一句话要点标签🔗⭐
44 Beyond Polarization: The Generative Constraint of Chain-of-Thought in Pointwise Reranking 提出链式思维模型以解决点对点重排序中的评分偏差问题 reinforcement learning chain-of-thought
45 WebWorld: The Browser as a World Model for Self-Improving Web Code 提出WebWorld以解决VLM驱动网页代码自我改进中的结构性缺陷 world model world models
46 Stratified Consistency Distillation for Natural Language Formalization 提出分层一致性蒸馏方法以提升自然语言到逻辑公式的翻译准确性 distillation large language model
47 COGTRL: Training LLMs for Scientific Discovery Assistance using Cognitive Traces via Reinforcement Learning 提出COGTRL以提升科学发现助手的认知能力 reinforcement learning large language model
48 SemPOI-RL: Aligning LLM Semantic Reasoning for Interpretable Out-of-Town POI Sequential Generation 提出SemPOI-RL以解决POI序列生成的可解释性问题 reinforcement learning masked autoencoder large language model ✅
49 GMTS: Gradient Magnitude-based Token Selection Improves RLVR Training for LLM Reasoning 提出GMTS方法以提升RLVR训练效果 reinforcement learning large language model
50 Mind the Gap: Theory-of-Mind-Grounded Friction for Epistemic Alignment 提出基于心智理论的摩擦优化以解决对话中的信念对齐问题 DPO reward shaping

🔬 支柱八：物理动画 (Physics-based Animation) (2 篇)

#题目一句话要点标签🔗⭐
51 Can LLMs Take the Pulse of the Economy? A Real-Time Evaluation of LLM Nowcasts on Macroeconomic Indicators 提出LiveMacroEval以实时评估大型语言模型的宏观经济指标预测能力 PULSE large language model
52 TRIPPULSE: Multi-Agent Travel Planning with Review-Grounded Reasoning 提出TRIPPULSE以解决基于评论的多代理旅行规划问题 PULSE

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
53 When Models Hear What They Expect: Diagnosing Prosodic Heuristics in Multimodal Sarcasm Detection 提出多模态模型的讽刺检测方法以解决音调启发式问题 manipulation large language model multimodal

⬅️ 返回 cs.CL 首页 · 🏠 返回主页