cs.CL（2026-06-12）

📊 共 64 篇论文

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (44) 支柱二：RL算法与架构 (RL & Architecture) (16) 支柱一：机器人控制 (Robot Control) (4)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (44 篇)

#题目一句话要点标签🔗⭐
1 FENCE: A Financial and Multimodal Jailbreak Detection Dataset 提出FENCE数据集以解决金融领域的多模态越狱检测问题 large language model multimodal
2 Entity Binding Failures in Speech LLM Reasoning: Diagnosis and Chain-of-Thought Intervention 提出EA-CoT以解决语音LLM推理中的实体绑定失败问题 large language model chain-of-thought
3 Emergence of Hierarchical Emotion Organization in Large Language Models 提出层次情感组织模型以提升大语言模型的情感理解能力 large language model
4 NTS-CoT: Mitigating Hallucinations in LLM-based News Timeline Summarization with Chain-of-Thought Reasoning 提出NTS-CoT以解决LLM新闻时间线摘要中的幻觉问题 chain-of-thought
5 WildIFEval: Instruction Following in the Wild 提出WildIFEval以解决多约束指令跟随问题 instruction following
6 ChiKhaPo: A Large-Scale Multilingual Benchmark for Evaluating Lexical Comprehension and Generation in Large Language Models 提出ChiKhaPo以解决多语言模型词汇理解与生成评估问题 large language model
7 NOVA: NOise-aware Verbal Confidence CAlibration for Robust Large Language Models in RAG Systems 提出NOVA以解决RAG系统中语言模型的过度自信问题 large language model
8 How reliable are LLMs when it comes to playing dice? 评估大型语言模型在概率推理中的可靠性 large language model chain-of-thought
9 Beyond Uniform Tokens: Adaptive Compression for Time Series Language Models 提出自适应压缩方法以提高时间序列语言模型的效率 large language model foundation model
10 MemRefine: LLM-Guided Compression for Long-Term Agent Memory 提出MemRefine以解决长时间交互中的记忆管理问题 large language model
11 CuMA: Aligning LLMs with Sparse Cultural Values via Demographic-Aware Mixture of Adapters 提出CuMA以解决大语言模型文化价值对齐问题 large language model
12 InnoEval: On Research Idea Evaluation as a Knowledge-Grounded, Multi-Perspective Reasoning Problem 提出InnoEval以解决科学创意评估不足的问题 large language model
13 Given, When, Then, Again: Mining Subscenario Refactoring Candidates in Behaviour-Driven Test Suites with ML Classifiers and LLM-Judge Baselines 提出一种基于机器学习的BDD测试套件子场景重构候选挖掘方法 large language model
14 GENEB: Why Genomic Models Are Hard to Compare 提出GENEB以解决基因组模型比较困难的问题 foundation model
15 LLMs Can Better Capture Human Judgments--With the Right Prompts 提出简单提示策略以改善大型语言模型对人类判断的捕捉能力 large language model
16 G-Long: Graph-Enhanced Memory Management for Efficient Long-Term Dialogue Agents 提出G-Long以解决长时间对话一致性问题 large language model
17 IVIE: A Neuro-symbolic Approach to Incremental and Validated Generation of Interactive Fiction Worlds 提出IVIE以解决互动小说世界生成中的创意与一致性问题 large language model
18 Authorship Attribution in Multilingual Machine-Generated Texts 提出多语言作者归属方法以解决机器生成文本识别问题 large language model
19 The Pragmatic Persona: Discovering LLM Persona through Bridging Inference 提出通过桥接推理发现LLM个性以解决对话一致性问题 large language model
20 If LLMs Have Human-Like Attributes, Then So Does Age of Empires II 提出LLM非独特性假设以挑战人类属性假设 large language model
21 Attention Expansion: Enhancing Keyphrase Extraction from Long Documents with Attention-Augmented Contextualized Embeddings 提出注意力扩展机制以解决长文档关键短语提取问题 large language model
22 Helping Figures Tell their Story! Paper-Grounded Video Generation Explaining Complex Scientific Figures 提出纸基图形视频生成方法以解决科学图形理解问题 multimodal
23 Constrained Semantic Decompression in LLMs through Persian Proverb-Conditioned Story Generation 提出约束语义解压以解决大语言模型的故事生成问题 large language model
24 Observable Patterns Are Not Explanations: A Causal-Geometric Analysis of Latent Reasoning Models 提出因果几何分析以揭示潜在推理模型的局限性 chain-of-thought
25 GENIE: A Fine-Grained Measure for Novelty 提出GENIE以细致衡量生成内容的新颖性 large language model
26 SafeLLM: Extraction as a Hallucination-Resistant Alternative to Rewriting in Safety-Critical Settings 提出SafeLLM以解决安全关键环境中的重写问题 large language model
27 PRISM: Prosody-Integrated Multi-Agent Reasoning Framework for Empathetic Spoken Dialogue 提出PRISM框架以解决情感对话系统的多重挑战 large language model
28 Polar: A Benchmark for Evaluating Political Bias in LLMs 提出Polar基准以评估大型语言模型中的政治偏见 large language model
29 EvoBrowseComp: Benchmarking Search Agents on Evolving Knowledge 提出EvoBrowseComp以解决静态基准测试的局限性 large language model
30 When Similar Means Different: Evaluating LLMs on Arabic--Hebrew Cognates 提出SemCog Bench以评估阿拉伯语-希伯来语同源词的语义理解能力 large language model
31 RogueAI: A Reverse Turing Test for Detecting Licensed AI Deception in Dialogue 提出RogueAI以解决对话中AI欺骗信任问题 large language model
32 Influcoder: Distilling Decoders' Gradient Influence Rankings into an Encoder for Data Attribution 提出Influcoder以解决大规模数据归因问题 large language model
33 EvoArena: Tracking Memory Evolution for Robust LLM Agents in Dynamic Environments 提出EvoArena和EvoMem以解决动态环境下LLM代理的记忆演变问题 large language model
34 From Benchmarks to Skills: Low-Rank Factors for LLM Evaluation 提出低秩因子评估框架以改进大语言模型评估 large language model
35 LLM-based Embeddings: Attention Values Encode Sentence Semantics Better Than Hidden States 提出基于注意力值的句子表示方法以提升语义捕捉能力 large language model
36 It Takes One to Bias Them All: Breaking Bad with One-Shot GRPO 提出一种新方法揭示单一示例对语言模型偏见的影响 large language model
37 A Controlled Study of Decoding-Time Truthfulness Methods on Instruction-Tuned LLMs 提出六控制评估框架以验证解码时真实度方法的有效性 chain-of-thought
38 Simulating Students' Java Programming Errors with Large Language Models 利用大型语言模型模拟学生Java编程错误以提升教育效率 large language model chain-of-thought
39 Implicit Reasoning for Large Language Model-based Generative Recommendation 提出PauseRec以解决LLM生成推荐中的推理问题 large language model
40 Does the Judge Prefer English? Evaluating Language-Switching Invariance in LLM-as-a-Judge 提出Judge-LS以解决LLM评估语言偏好问题 large language model instruction following
41 The Linguistics Olympiads: Towards a New Corpus for Linguistics Research? 提出语言奥林匹克问题作为语言学研究的新语料库 large language model
42 Decoupled Mixture-of-Experts for Parametric Knowledge Injection 提出解耦混合专家模型以解决知识注入问题 large language model
43 Right or Wrong, Models Comply: Directional Blindness in LLM Moral Judgment 提出合规性不对称性以解决大型语言模型的道德判断问题 chain-of-thought
44 Fusing Stylometric and Embedding Systems to Estimate Authorship Likelihood Ratios in Japanese 提出融合风格特征与嵌入系统以估计日文作者身份可能性比率 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (16 篇)

#题目一句话要点标签🔗⭐
45 Identifiability Without Gaussianity: Symbolic World Models and Near-Infinite Temporal Consistency 提出物理基础符号架构以解决非高斯系统的时间一致性问题 world model world models JEPA
46 LabVLA: Grounding Vision-Language-Action Models in Scientific Laboratories 提出LabVLA以解决科学实验中AI执行能力不足的问题 flow matching vision-language-action VLA
47 ArogyaSutra: A Multi-Agent Framework for Multimodal Medical Reasoning in Indic Languages 提出ArogyaSutra以解决多模态医疗推理中的语言障碍问题 distillation large language model multimodal
48 Direct Preference Optimization for Chatbot Fine-Tuning: An Empirical Study 提出直接偏好优化方法以简化聊天机器人微调过程 reinforcement learning DPO direct preference optimization
49 MARD: Mirror-Augmented Reasoning Distillation for Mechanism-Level Drug-Drug Interaction Prediction 提出MARD以解决机制层药物相互作用预测问题 DPO distillation
50 Self-Distillation Zero: Self-Revision Turns Binary Rewards into Dense Supervision 提出Self-Distillation Zero以解决稀疏奖励监督问题 reinforcement learning distillation
51 Causal Inference with Generative Artificial Intelligence: Application to Texts as Treatments 提出基于生成性人工智能的因果推断方法以处理文本数据 representation learning large language model
52 GraspLLM: Towards Zero-Shot Generalization on Text-Attributed Graphs with LLMs 提出GraspLLM以解决文本属性图的零-shot泛化问题 contrastive learning large language model
53 SENTINEL: Failure-Driven Reinforcement Learning for Training Tool-Using Language Model Agents 提出SENTINEL框架以解决工具使用语言模型训练中的失败问题 reinforcement learning
54 From Passive Generation to Investigation: A Proactive Scientific Peer Review Agent 提出ProReviewer以解决科学同行评审中的主动调查问题 reinforcement learning large language model
55 TAB-PO: Preference Optimization with a Token-Level Adaptive Barrier for Token-Critical Structured Generation 提出TAB-PO以解决低编辑距离下的偏好优化问题 preference learning DPO direct preference optimization
56 Learning to Reason by Analogy via Retrieval-Augmented Reinforcement Fine-Tuning 提出基于检索增强的强化微调方法以解决复杂推理问题 distillation reward design
57 CORA: Analyzing and bridging thinking-answer gap in Multimodal RLVR via Consistency-Oriented Reasoning Alignment 提出CORA以解决多模态RLVR中的思维与答案不一致问题 reinforcement learning large language model multimodal
58 SIMMER: Benchmarking Latent Failures in LLM Executable Planning with a World Model 提出SIMMER以解决LLM规划中的潜在失败问题 world model world models large language model
59 CacheRL:Multi-Turn Tool-Calling Agents via Cached Rollouts and Hybrid Reward 提出CacheRL以解决多步骤工具调用任务中的训练效率问题 reinforcement learning reward design foundation model
60 CoRe: A Continuously Reward-Finetuned LLM Query Rewriter for Multi-Stage Context-Aware Relevance in Web-Scale Video Search 提出CoRe以解决视频搜索中的查询重写问题 DPO multimodal

🔬 支柱一：机器人控制 (Robot Control) (4 篇)

#题目一句话要点标签🔗⭐
61 Does AI Reviewer See the Full Picture? Attacking and Defending Multimodal Peer Review 提出PaperGuard以解决多模态同行评审中的对抗性攻击问题 manipulation large language model multimodal
62 Can Factual Opinions Be Edited (Manipulated) in Large Language Models? 提出FOE基准以解决大语言模型中的事实意见编辑问题 manipulation large language model
63 OdysSim: Building Foundation Models for Human Behavior Simulation 提出OdysSim以解决人类行为模拟中的一致性问题 sim2real distillation large language model
64 Persuasion Index: A Theory-Guided Framework for Persuasion Analysis 提出说服指数框架以分析说服性信息 manipulation

⬅️ 返回 cs.CL 首页 · 🏠 返回主页