cs.CL（2026-04-08）

📊 共 32 篇论文 | 🔗 1 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (29 🔗1) 支柱一：机器人控制 (Robot Control) (1) 支柱二：RL算法与架构 (RL & Architecture) (1) 支柱七：动作重定向 (Motion Retargeting) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (29 篇)

#题目一句话要点标签🔗⭐
1 To Adapt or not to Adapt, Rethinking the Value of Medical Knowledge-Aware Large Language Models 重新评估医学知识增强大语言模型的价值，揭示其在特定场景下的局限性与潜力 large language model instruction following
2 Multilingual Cognitive Impairment Detection in the Era of Foundation Models 利用预训练模型和语言特征进行多语种认知障碍检测 large language model foundation model
3 The Impact of Steering Large Language Models with Persona Vectors in Educational Applications 研究人格向量引导大型语言模型在教育应用中的影响，揭示任务和架构敏感性。 large language model
4 Self-Preference Bias in Rubric-Based Evaluation of Large Language Models 揭示基于准则评估中大语言模型的自我偏好偏差问题 large language model
5 LaScA: Language-Conditioned Scalable Modelling of Affective Dynamics LaScA：提出一种基于语言条件的可扩展情感动态建模方法，提升情感预测的准确性和可解释性。 language conditioned
6 On the Step Length Confounding in LLM Reasoning Data Selection 揭示并缓解LLM推理数据选择中步长偏差问题，提升数据质量 large language model chain-of-thought
7 Luwen Technical Report 提出Luwen：一个基于Baichuan的中文法律大语言模型，提升法律领域任务性能。 large language model foundation model
8 iTAG: Inverse Design for Natural Text Generation with Accurate Causal Graph Annotations 提出iTAG，通过逆向设计和精确因果图标注生成自然文本，解决文本因果发现中数据匮乏问题。 large language model chain-of-thought
9 GCoT-Decoding: Unlocking Deep Reasoning Paths for Universal Question Answering 提出GCoT-Decoding，解锁通用问答任务的深度推理路径 large language model chain-of-thought
10 DiffuMask: Diffusion Language Model for Token-level Prompt Pruning DiffuMask：提出基于扩散语言模型的token级别prompt并行剪枝方法 large language model chain-of-thought
11 Dynamic Context Evolution for Scalable Synthetic Data Generation 提出动态上下文演化（DCE）框架，解决大规模合成数据生成中的跨批次模式崩溃问题。 large language model
12 Adaptive Prompt Structure Factorization: A Framework for Self-Discovering and Optimizing Compositional Prompt Programs 提出自适应提示结构分解框架，用于自动发现和优化组合提示程序。 large language model
13 The Detection--Extraction Gap: Models Know the Answer Before They Can Say It 提出黑箱自适应早期退出以解决检测与提取之间的差距问题 chain-of-thought ✅
14 LLM-based Schema-Guided Extraction and Validation of Missing-Person Intelligence from Heterogeneous Data Sources 提出Guardian Parser Pack，利用LLM从异构数据源中提取和验证失踪人员情报 large language model
15 The Illusion of Stochasticity in LLMs 大型语言模型在随机采样方面存在缺陷，影响其作为智能体的可靠性 large language model
16 The AI Skills Shift: Mapping Skill Obsolescence, Emergence, and Transition Pathways in the LLM Era 提出SAFI评估LLM对职业技能的影响，揭示技能自动化可行性与转型路径。 large language model
17 WRAP++: Web discoveRy Amplified Pretraining WRAP++：通过Web关系发现增强预训练，提升LLM知识获取 large language model
18 TeamLLM: A Human-Like Team-Oriented Collaboration Framework for Multi-Step Contextualized Tasks 提出TeamLLM框架，模拟人类团队协作解决多步骤上下文任务 large language model
19 Why teaching resists automation in an AI-inundated era: Human judgment, non-modular work, and the limits of delegation 论AI时代教学为何难以自动化：人的判断、非模块化工作及委托的局限性 large language model
20 Agent-Driven Corpus Linguistics: A Framework for Autonomous Linguistic Discovery 提出Agent-Driven Corpus Linguistics框架，实现基于LLM的自主语言发现。 large language model
21 Language Bias under Conflicting Information in Multilingual LLMs 揭示多语言LLM在冲突信息处理中存在的语言偏见，尤其对俄语和中文存在显著倾向。 large language model
22 Gemma 4, Phi-4, and Qwen3: Accuracy-Efficiency Tradeoffs in Dense and MoE Reasoning Language Models 对比Gemma、Phi和Qwen3，评估稠密模型与MoE模型在推理任务中的精度-效率权衡。 chain-of-thought
23 Corpora deduplication or duplication in Natural Language Processing of few resourced languages ? A case of study: The Mexico's Nahuatl 针对低资源语言，研究语料重复对自然语言处理的影响：以墨西哥纳瓦特尔语为例 large language model
24 AGSC: Adaptive Granularity and Semantic Clustering for Uncertainty Quantification in Long-text Generation 提出AGSC框架，通过自适应粒度和语义聚类提升长文本生成中不确定性量化的准确性和效率。 large language model
25 Cognitive Loop of Thought: Reversible Hierarchical Markov Chain for Efficient Mathematical Reasoning 提出认知回路思维（CLoT），解决LLM数学推理中长序列和上下文丢失问题。 chain-of-thought
26 From Perception to Autonomous Computational Modeling: A Multi-Agent Approach 提出基于多智能体LLM的计算建模框架，实现从感知数据到自主工程报告的完整流程。 large language model
27 When Is Thinking Enough? Early Exit via Sufficiency Assessment for Efficient Reasoning 提出DTSR框架，通过充分性评估实现大语言模型高效推理的提前退出。 chain-of-thought
28 StructKV: Preserving the Structural Skeleton for Scalable Long-Context Inference StructKV：通过保留结构骨架实现可扩展的长上下文推理 large language model
29 Argus: Reorchestrating Static Analysis via a Multi-Agent Ensemble for Full-Chain Security Vulnerability Detection Argus：通过多智能体集成重构静态分析，实现全链安全漏洞检测 large language model

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
30 Continuous Interpretive Steering for Scalar Diversity 提出连续解释性引导(CIS)方法，评估LLM中标量多样性的语用推理能力。 manipulation large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (1 篇)

#题目一句话要点标签🔗⭐
31 Personalized RewardBench: Evaluating Reward Models with Human Aligned Personalization 提出 Personalized RewardBench，用于评估奖励模型对个性化偏好的建模能力 PPO large language model

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#题目一句话要点标签🔗⭐
32 OpenSpatial: A Principled Data Engine for Empowering Spatial Intelligence 提出OpenSpatial数据引擎，赋能空间智能，并构建3百万规模高质量数据集。 spatial relationship

⬅️ 返回 cs.CL 首页 · 🏠 返回主页