cs.CL（2026-06-04）

📊 共 48 篇论文 | 🔗 6 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (34 🔗5) 支柱二：RL算法与架构 (RL & Architecture) (12 🔗1) 支柱一：机器人控制 (Robot Control) (2)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (34 篇)

#题目一句话要点标签🔗⭐
1 Harnessing Structural Context for Entity Alignment Foundation Models 提出ContextEA以解决知识图谱实体对齐中的结构上下文不足问题 foundation model
2 LLMs Can Leak Training Data But Do They Want To? A Propensity-Aware Evaluation of Memorization in LLMs 提出PropMe框架以评估大型语言模型的记忆能力 large language model
3 Evaluating Stochastic Collapse and Implicit Bias in Multimodal Large Language Models 提出RandomBench以评估多模态大语言模型的随机性与隐含偏差问题 large language model multimodal
4 Mechanistic Insights into Functional Sparsity in Multimodal LLMs via CoRe Heads 提出CoRe头以揭示多模态大语言模型中的功能稀疏性 large language model multimodal
5 To Be Multimodal or Not to Be: Query-Adaptive Audio-Visual Person Retrieval via Active Modality Detection 提出查询自适应框架以解决多模态人检索问题 multimodal
6 IR3DE: A Linear Router for Large Language Models 提出IR3DE以解决大型语言模型的高效路由问题 large language model
7 MARDoc: A Memory-Aware Refinement Agent Framework for Multimodal Long Document QA 提出MARDoc框架以解决长文档多模态问答中的信息稀疏问题 multimodal
8 The Tell-Tale Norm: $\ell_2$ Magnitude as a Signal for Reasoning Dynamics in Large Language Models 提出l2范数作为大型语言模型推理动态的信号 large language model ✅
9 Large Language Models are Perplexed by some Political Parties 评估大型语言模型在政治公平性上的表现 large language model
10 Analysis of the Neglect-Zero Effect in Large Language Models 探讨大语言模型中的忽视零效应及其认知过程 large language model ✅
11 An Embarrassingly Simple Detector for Model Extraction Attacks in Large Language Model API Traffic 提出一种简单有效的检测器以应对大语言模型API的模型提取攻击 large language model ✅
12 AdaPlanBench: Evaluating Adaptive Planning in Large Language Model Agents under World and User Constraints 提出AdaPlanBench以解决大语言模型在动态约束下的自适应规划问题 large language model
13 Using Large Language Models to Support High Volume Application Review for an Undergraduate Research Program 基于大型语言模型的工具助力本科研究项目申请评审 large language model
14 Latent Reasoning with Normalizing Flows 提出NF-CoT框架以提升潜在推理能力 large language model chain-of-thought
15 Revising Context, Shifting Simulated Stance: Auditing LLM-Based Stance Simulation in Online Discussions 提出基于反事实上下文修订的框架以审计LLM立场模拟 large language model multimodal
16 IA-RAG: Interval-Algebra-Driven Temporal Reasoning for Dynamic Knowledge Retrieval 提出IA-RAG框架以解决动态知识检索中的时间推理问题 large language model TAMP ✅
17 Beyond tokens: a unified framework for latent communication in LLM-based multi-agent systems 提出统一框架以解决LLM多智能体系统中的潜在通信问题 large language model chain-of-thought
18 When New Generators Arrive: Lifelong Machine-Generated Text Attribution via Ridge Feature Transfer 提出RidgeFT以解决长期机器生成文本归属问题 large language model
19 Re-Centering Humans in LLM Personalization 提出人类数据驱动的LLM个性化评估方法以解决现有系统局限性 large language model
20 Human Adults and LLMs as Scientists: Who Benefits from Active Exploration? 通过主动探索提升成人的因果推理能力 large language model
21 Value-and-Structure Alignment for Routing-Consistent Quantization of Mixture-of-Experts Models 提出VSRAQ以解决MoE模型量化中的路由不一致问题 foundation model
22 PromptPrint: Behavioral Biometrics Through Natural Language Prompting in LLMs 提出PromptPrint以解决用户身份识别问题 large language model
23 When to Think Deeply: Inhibitory Deliberation for LLM Reasoning 提出IDPR框架以优化LLM推理效率 large language model
24 UnpredictaBench: A Benchmark for Evaluating Distributional Randomness in LLMs 提出UnpredictaBench以评估LLMs的分布随机性 large language model
25 Scaffold, Not Vocabulary? A Controlled, Two-Tier, Pre-Registered Study of a Popperian Code-Generation Skill 提出两层次的消融研究以评估代码生成技能的有效性 large language model
26 FOXGLOVE: Understanding Goal-Oriented and Anchored Writing Feedback from Experts and LLMs on Argumentative Essays 提出FOXGLOVE以系统比较专家与LLM在写作反馈中的差异 large language model
27 Automatic Labelling of Speech Translation Errors 提出语音翻译错误自动标注方法以提升系统可信度 multimodal
28 Contextualized Prompting For Stance Detection On Social Media 提出上下文化提示以解决社交媒体立场检测问题 large language model ✅
29 The Generator-Eraser Paradox: Community Guidelines for Responsible LLM-Assisted Dialect Resource Creation 提出生成器-消除者悖论以指导负责任的方言资源创建 large language model
30 ReverseEOL: Improving Training-free Text Embeddings via Text Reversal in Decoder-only LLMs 提出ReverseEOL以提升无训练文本嵌入的表示能力 large language model
31 ProSPy: A Profiling-Driven SQL-Python Agentic Framework for Enterprise Text-to-SQL 提出ProSPy框架以解决企业级Text-to-SQL的挑战 large language model
32 Can LLMs Be Constrained to the Past? Improving Knowledge Cutoff through Recall-Based Prompting 提出基于回忆的提示策略以改善知识截止问题 large language model
33 PlanBench-V: A Spatial Planning Map Benchmark for Vision-Language Models 提出PlanBench-V以解决空间规划图解释的评估问题 multimodal
34 Predictable Scaling Laws of Optimal Hyperparameters for LLM Continued Pre-training 提出可预测的超参数缩放法则以优化大语言模型继续预训练 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (12 篇)

#题目一句话要点标签🔗⭐
35 EGTR-Review: Efficient Evidence-Grounded Scientific Peer Review Generation via Multi-Agent Teacher Distillation 提出EGTR-Review以解决科学同行评审生成中的证据支持不足问题 distillation large language model
36 Emergent Language as an Approach to Conscious AI 提出基于新兴语言的方法以研究意识AI reinforcement learning affordance
37 USAD 2.0: Scaling Representation Distillation for Universal Audio Understanding 提出USAD 2.0以解决音频理解中的多域编码问题 distillation large language model foundation model
38 Reinforcement Learning Elicits Contextual Learning of Unseen Language Translation 提出强化学习方法以解决低资源语言翻译问题 reinforcement learning large language model zero-shot transfer
39 Improving Cross-Lingual Factual Recall via Consistency-Driven Reinforcement Learning 提出PolyFact以解决跨语言事实一致性问题 reinforcement learning large language model
40 TARPO: Token-Wise Latent-Explicit Reasoning via Action-Routing Policy Optimization 提出TARPO以解决强化学习中的策略探索问题 reinforcement learning large language model chain-of-thought ✅
41 What Do People Actually Want From AI? Mapping Preference Plurality 揭示AI偏好多样性以改善人机对齐方法 reinforcement learning RLHF large language model
42 Data-Efficient Autoregressive-to-Diffusion Language Models via On-Policy Distillation 提出基于策略蒸馏的自回归到扩散语言模型转换方法 distillation
43 Interpreting Style Representations via Style-Eliciting Prompts 提出风格引导提示以解决风格表示解释问题 representation learning large language model
44 CHASE: Adversarial Red-Blue Teaming for Improving LLM Safety using Reinforcement Learning 提出CHASE框架以提升大型语言模型的安全性 reinforcement learning
45 Better Literary Translation: A Multi-Aspect Data Generation and LLM Training Approach 提出多维数据生成与LLM训练方法以提升文学翻译质量 reinforcement learning DPO
46 EDIT: Evidence-Diagnosed Intervention Training for Rule-Faithful LLM Grading 提出EDIT框架以解决LLM评分的可靠性问题 reinforcement learning reward shaping

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
47 CollabSim: A CSCW-Grounded Methodology for Investigating Collaborative Competence of LLM Agents through Controlled Multi-Agent Experiments 提出CollabSim以系统评估多代理系统的协作能力 manipulation large language model
48 Decomposing Factual Sycophancy in Language Models: How Size and Instruction Tuning Shape Robustness 提出分解语言模型中的事实谄媚以提升鲁棒性 manipulation

⬅️ 返回 cs.CL 首页 · 🏠 返回主页