cs.CL（2026-09-30）

📊 共 33 篇论文 | 🔗 4 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (22 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (10 🔗3) 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 Beyond Text: LLM-Based Dimensional Emotion Evaluation in Multimodal Dialogue 提出基于LLM的框架以解决多模态对话中的情感评估问题 large language model multimodal
2 Structure vs. Chain-of-Thought: Evaluating LLM Criteria Extraction for Depression Severity 比较结构与思维链方法在抑郁症严重程度评估中的应用 large language model chain-of-thought
3 On the (In)effectiveness of AMR Augmentation for Large Language Models 探讨AMR增强对大型语言模型的有效性问题 large language model
4 Cognitive Enhancement: Rethinking the Necessity of Role-Playing for Large Language Models 提出MLCP以解决角色扮演对LLM性能影响不确定性问题 large language model
5 The Evolution of Attention in Large Language Models: Mechanisms, Trade-offs, and Emerging Trends 提出五维视角分析大语言模型中的注意力机制演变 large language model
6 QuantCode Model: Specializing Language Models for Executable Algorithmic Trading Code 提出QuantCode模型以解决可执行算法交易代码生成问题 large language model instruction following
7 Making LLMs Say What They Think: Measuring and Improving CoT-Interpretability Alignment 提出CoT-可解释性对齐度量以提升LLM推理透明度 large language model chain-of-thought ✅
8 LLM Persona Unlearning 提出PersonaUnlearnBench以解决大语言模型的角色遗忘问题 large language model
9 RAIM: Robust Aggregation of Inexpensive Models for Hallucination Detection 提出RAIM以实现低成本模型的聚合用于幻觉检测 large language model
10 EvoDuet: Bilevel Co-Evolution of Web Searching and Task Solving for Scientific Discovery 提出EvoDuet以解决科学发现中的知识缺口问题 large language model
11 Decision-Oriented Recommendation Reranking: An Empirical Study of Jev 提出决策导向推荐重排序模型Jev以解决推荐效率与质量的权衡问题 large language model
12 Index-Translate: A Multilingual Translation Model Family -- Text, Speech, Controlled Dubbing, and Long-Document Translation 提出Index-Translate以解决多语言翻译任务的多样性问题 instruction following
13 Debias It Yourself: Teaching LLMs Cognitive Bias Mitigation Interventions 提出Debias It Yourself框架以缓解大型语言模型的认知偏见 large language model
14 JuryFlow: Disagreement-Guided Human-in-the-Loop Multi-Agent Evaluation 提出JuryFlow框架以解决评估中评审者分歧问题 large language model
15 The Concrete-Arbitrary Gap: Kinship Reasoning in LLMs Is Not Indifferent to Presentation 研究表明语言模型在亲属关系推理中对表达方式敏感 large language model
16 Drift Inspector: Exploring and Measuring Scientific Drift with Atomic Contribution Claims 提出Drift Inspector以测量科学领域的变化 multimodal
17 Marginal Response Surface Elicitation for Zero-Label Tabular Learning 提出MARS方法以解决零标签表格学习问题 large language model
18 Argument Structure Prediction in Online Conversations: A Comparative Study of Modeling Paradigms and Task Architectures 提出对话中论证结构预测的比较研究以解决建模挑战 large language model
19 RSIGame: Autonomous Agentic Game Development with Recursive Self-improvement 提出RSIGame框架以解决自动游戏生成中的质量提升问题 large language model
20 You're Hired: Strategic Model Selection for LLM Collaboration 提出多模型选择算法以优化LLM协作系统 large language model
21 Uncovering Uncontrolled Repetition through Residual Stream Dynamics 提出Tokenwise Residual Comparison以解决大语言模型中的无控制重复问题 large language model
22 MetaSteer: Context-Conditioned, nonlinear Steering via Attention-Projection Adaptation 提出MetaSteer以解决大语言模型的非线性引导问题 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (10 篇)

#题目一句话要点标签🔗⭐
23 StateTree: Enhancing Long-Term Dialogue Reasoning via Reinforcement Learning 提出StateTree以解决长时间对话推理中的数据稀缺问题 reinforcement learning large language model TAMP
24 CATCH: A Controllable Analysis Testbed for Reward Hacking in Coding RL 提出CATCH以解决强化学习中的奖励黑客问题 reinforcement learning reward design large language model ✅
25 OPSRD: On-Policy Self-Role Distillation 提出OPSRD以解决角色提示下的自我蒸馏问题 distillation large language model ✅
26 Diagnosing On-Policy Self-Distillation for Reasoning Language Models 诊断在政策自蒸馏以提升语言模型推理能力 distillation privileged information
27 When Clipping Reverses Correction: Failure Dynamics of Pointwise Forward-KL On-Policy Self-Distillation 揭示点对点前向KL剪切对自蒸馏训练的负面影响 distillation privileged information
28 Better Supervision Is Nearby: Neighborhood On-Policy Self-Distillation 提出邻域在线自蒸馏方法以提升数学推理模型性能 distillation
29 LatentHarness: Learning Latent Actions for Memory and Reasoning via Counterfactual Policy Distillation 提出LatentHarness以解决长上下文推理中的记忆与推理瓶颈问题 distillation
30 Offline Guidance, Online Reasoning: Reusing LLM Feedback for Small Language Models 提出可重用潜在修正方法以提升小型语言模型推理能力 distillation large language model ✅
31 LexReward: A Taxonomy-Driven Reward Framework for Legal Language Models 提出LexReward框架以提升法律语言模型的奖励信号质量 reinforcement learning DPO direct preference optimization
32 Training LLM Judges from Language Feedback via Position-Selective Self-Distillation 通过位置选择自蒸馏训练LLM评判者以优化语言反馈 distillation

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
33 Spike-driven Vision-Language-Action Model 提出Spike驱动的视觉-语言-动作模型以解决资源受限平台的控制问题 manipulation scene understanding vision-language-action

⬅️ 返回 cs.CL 首页 · 🏠 返回主页