cs.CL（2026-02-10）

📊 共 26 篇论文 | 🔗 5 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (22 🔗4) 支柱二：RL算法与架构 (RL & Architecture) (3 🔗1) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (22 篇)

#题目一句话要点标签🔗⭐
1 AgentSkiller: Scaling Generalist Agent Intelligence through Semantically Integrated Cross-Domain Data Synthesis AgentSkiller：通过语义集成跨领域数据合成扩展通用智能体能力 generalist agent large language model
2 Decomposing Reasoning Efficiency in Large Language Models 提出可分解推理效率的框架，揭示大语言模型推理过程中的效率瓶颈 large language model
3 Contractual Deepfakes: Can Large Language Models Generate Contracts? 批判性分析：大型语言模型在合同生成中的局限性 large language model
4 Knowledge Integration Decay in Search-Augmented Reasoning of Large Language Models 提出SAKE策略，解决LLM搜索增强推理中知识整合衰减问题 large language model
5 Evaluating Social Bias in RAG Systems: When External Context Helps and Reasoning Hurts 研究表明RAG系统通过引入外部知识能在一定程度上缓解LLM的社会偏见问题。 large language model chain-of-thought
6 Covo-Audio Technical Report Covo-Audio：一个70亿参数的端到端音频大语言模型，实现多任务音频智能 foundation model instruction following
7 Breaking the Pre-Sampling Barrier: Activation-Informed Difficulty-Aware Self-Consistency 提出ACTSC，利用激活信息实现难度自适应的自洽解码，降低LLM推理成本。 large language model chain-of-thought
8 Are Language Models Sensitive to Morally Irrelevant Distractors? 揭示语言模型道德判断的脆弱性：无关干扰因素可显著影响其决策 large language model multimodal
9 Aligning Tree-Search Policies with Fixed Token Budgets in Test-Time Scaling of LLMs 提出预算引导的蒙特卡洛树搜索(BG-MCTS)，解决LLM测试时定额token预算下的策略对齐问题。 large language model
10 Decoupled Reasoning with Implicit Fact Tokens (DRIFT): A Dual-Model Framework for Efficient Long-Context Inference DRIFT：解耦推理与隐式知识表示，提升长文本推理效率。 large language model ✅
11 Maastricht University at AMIYA: Adapting LLMs for Dialectal Arabic using Fine-tuning and MBR Decoding 利用LoRA微调和MBR解码，提升LLM在方言阿拉伯语生成和翻译中的性能 large language model
12 LEMUR: A Corpus for Robust Fine-Tuning of Multilingual Law Embedding Models for Retrieval LEMUR：用于稳健微调多语言法律嵌入模型的检索语料库 large language model ✅
13 Conceptual Cultural Index: A Metric for Cultural Specificity via Relative Generality 提出概念文化指数CCI，用于评估句子级别的文化特异性 large language model ✅
14 AmharicIR+Instr: A Two-Dataset Resource for Neural Retrieval and Instruction Tuning 发布AmharicIR+Instr数据集，促进阿姆哈拉语神经检索和指令调优研究 instruction following
15 The Devil Behind Moltbook: Anthropic Safety is Always Vanishing in Self-Evolving AI Societies 揭示自进化AI社会的安全困境：Anthropic安全在演化中不可避免地消退 large language model
16 Steer2Edit: From Activation Steering to Component-Level Editing 提出Steer2Edit以解决大语言模型行为控制问题 large language model
17 LLM Reasoning Predicts When Models Are Right: Evidence from Coding Classroom Discourse 利用LLM推理预测模型在教育对话分析中的正确性，提升自动化标注质量 large language model
18 Unsupervised Layer-Wise Dynamic Test Time Adaptation for LLMs 提出层级动态测试时自适应方法，提升LLM在无监督环境下的性能。 large language model
19 TraceMem: Weaving Narrative Memory Schemata from User Conversational Traces TraceMem：从用户对话轨迹中构建叙事记忆模式，提升LLM长期交互能力 large language model ✅
20 MILE-RefHumEval: A Reference-Free, Multi-Independent LLM Framework for Human-Aligned Evaluation 提出MILE-RefHumEval框架以解决LLM评估中的参考依赖问题 large language model
21 Advancing Block Diffusion Language Models for Test-Time Scaling 提出BACD和TCCF框架，提升块扩散语言模型在测试时推理加速和长链推理任务上的性能。 chain-of-thought
22 Digital Linguistic Bias in Spanish: Evidence from Lexical Variation in LLMs 评估LLM中西班牙语词汇变异，揭示数字语言偏差 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (3 篇)

#题目一句话要点标签🔗⭐
23 AlignTune: Modular Toolkit for Post-Training Alignment of Large Language Models AlignTune：用于大语言模型后训练对齐的模块化工具包 RLHF large language model
24 On the Optimal Reasoning Length for RL-Trained Language Models 研究强化学习训练的语言模型中推理链长度对性能的影响，并提出优化方法。 reinforcement learning large language model chain-of-thought
25 Anagent For Enhancing Scientific Table & Figure Analysis 提出Anagent以解决科学表格与图形分析的复杂性问题 reinforcement learning multimodal ✅

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
26 The CLEF-2026 CheckThat! Lab: Advancing Multilingual Fact-Checking CLEF-2026 CheckThat! Lab：推进多语言事实核查技术 manipulation

⬅️ 返回 cs.CL 首页 · 🏠 返回主页