cs.CL（2025-02-10）

📊 共 45 篇论文 | 🔗 9 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (36 🔗7) 支柱二：RL算法与架构 (RL & Architecture) (9 🔗2)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (36 篇)

#题目一句话要点标签🔗⭐
1 From No to Know: Taxonomy, Challenges, and Opportunities for Negation Understanding in Multimodal Foundation Models 提出多模态否定理解分类法，应对多模态大模型在否定语义理解上的挑战。 foundation model multimodal
2 Large Language Models Meet Symbolic Provers for Logical Reasoning Evaluation 提出ProverGen框架，结合LLM与符号证明器生成高质量一阶逻辑推理数据集ProverQA。 large language model chain-of-thought ✅
3 Structural Reformation of Large Language Model Neuron Encapsulation for Divergent Information Aggregation 提出结构化神经元封装，提升大语言模型信息聚合与逻辑推理能力 large language model
4 Multi-turn Evaluation of Anthropomorphic Behaviours in Large Language Models 提出多轮评估方法，用于衡量大型语言模型中拟人化行为的程度。 large language model
5 Specializing Large Language Models to Simulate Survey Response Distributions for Global Populations 提出一种基于微调LLM的方法，用于模拟全球人口的调查响应分布。 large language model
6 Demystifying Singular Defects in Large Language Models 揭示大语言模型奇异缺陷：基于奇异向量分析高范数Token现象 large language model ✅
7 Boosting Self-Efficacy and Performance of Large Language Models via Verbal Efficacy Stimulations 通过语言效能刺激提升大型语言模型的自我效能与表现 large language model
8 Hephaestus: Improving Fundamental Agent Capabilities of Large Language Models through Continual Pre-Training Hephaestus：通过持续预训练提升大语言模型智能体的基础能力 large language model
9 A Survey of Theory of Mind in Large Language Models: Evaluations, Representations, and Safety Risks 综述大型语言模型中的心理理论：评估、表征与安全风险 large language model
10 Systematic Outliers in Large Language Models 深入分析LLM中的系统性异常值，揭示其成因、功能及对模型的影响 large language model ✅
11 Latent Convergence Modulation in Large Language Models: A Novel Approach to Iterative Contextual Realignment 提出潜在收敛调制方法，提升大型语言模型长文本生成中的上下文一致性。 large language model
12 DebateBench: A Challenging Long Context Reasoning Benchmark For Large Language Models 提出 DebateBench：一个用于评估大型语言模型长文本推理能力的挑战性基准 large language model
13 GuideLLM: Exploring LLM-Guided Conversation with Applications in Autobiography Interviewing 提出GuideLLM，探索LLM引导的对话在自传访谈中的应用 large language model instruction following
14 Non-literal Understanding of Number Words by Language Models 通过链式思考提示，提升大语言模型对数字词汇的非字面理解能力 large language model chain-of-thought
15 ConMeC: A Dataset for Metonymy Resolution with Common Nouns ConMeC：一个用于普通名词转喻消解的数据集 large language model chain-of-thought ✅
16 Cardiverse: Harnessing LLMs for Novel Card Game Prototyping Cardiverse：利用大型语言模型进行创新卡牌游戏原型设计 large language model ✅
17 Tokenization Standards for Linguistic Integrity: Turkish as a Benchmark 提出一种新框架以评估土耳其语的分词策略 large language model
18 AIMS.au: A Dataset for the Analysis of Modern Slavery Countermeasures in Corporate Statements 提出AIMS.au数据集，用于分析企业声明中现代奴隶制应对措施 large language model
19 Finding Words Associated with DIF: Predicting Differential Item Functioning using LLMs and Explainable AI 利用LLM和可解释AI预测DIF，发现与DIF相关的词汇以提升评估公平性 large language model
20 Investigating the Zone of Proximal Development of Language Models for In-Context Learning 利用近端发展区理论分析LLM的上下文学习能力，提升推理和微调效果 large language model
21 Can 1B LLM Surpass 405B LLM? Rethinking Compute-Optimal Test-Time Scaling 提出计算最优的测试时缩放策略，使小模型在复杂任务上超越大模型 large language model
22 In-Context Learning (and Unlearning) of Length Biases 研究表明大语言模型能通过上下文学习长度偏差，并可用于消除模型自身编码的长度偏差。 large language model
23 Transparent NLP: Using RAG and LLM Alignment for Privacy Q&A 提出MultiRAIN对齐RAG系统，提升隐私问答中LLM的透明性和合规性 large language model
24 Do we really have to filter out random noise in pre-training data for language models? 研究表明预训练数据中的随机噪声对语言模型影响有限，并提出局部梯度匹配损失。 multimodal
25 LawGPT: Knowledge-Guided Data Generation and Its Application to Legal LLM 提出知识引导的数据生成框架KgDG，提升开源法律LLM的推理能力 large language model ✅
26 Adaptive Prompting: Ad-hoc Prompt Composition for Social Bias Detection 提出自适应Prompt组合方法，用于提升社交偏见检测任务的性能。 large language model
27 KARMA: Leveraging Multi-Agent LLMs for Automated Knowledge Graph Enrichment KARMA：利用多智能体LLM自动丰富知识图谱 large language model
28 Can AI Examine Novelty of Patents?: Novelty Evaluation Based on the Correspondence between Patent Claim and Prior Art 提出基于LLM的专利新颖性评估方法，并构建首个相关数据集。 large language model
29 SeaExam and SeaBench: Benchmarking LLMs with Local Multilingual Questions in Southeast Asia 提出SeaExam和SeaBench，用于评估LLM在东南亚本地多语言场景下的能力。 large language model
30 Krutrim LLM: Multilingual Foundational Model for over a Billion People Krutrim LLM：为十亿人口设计的印度语多语言基础模型 foundation model
31 Jakiro: Boosting Speculative Decoding with Decoupled Multi-Head via MoE Jakiro：利用MoE解耦多头注意力机制，加速推测解码并提升精度。 large language model ✅
32 Emergent Response Planning in LLMs 揭示LLM涌现的响应规划能力：隐藏层编码未来输出属性 large language model
33 Is LLM an Overconfident Judge? Unveiling the Capabilities of LLMs in Detecting Offensive Language with Annotation Disagreement 揭示LLM在处理标注不一致的冒犯性语言检测中的能力与过度自信问题 large language model
34 Scaling Public Health Text Annotation: Zero-Shot Learning vs. Crowdsourcing for Improved Efficiency and Labeling Accuracy 探索LLM零样本学习在公共健康文本标注中的应用，提升效率并评估标注准确性。 large language model
35 LegalViz: Legal Text Visualization by Text To Diagram Generation 提出LegalViz数据集，用于法律文本到易理解图表的生成，提升法律知识可访问性。 large language model
36 LCIRC: A Recurrent Compression Approach for Efficient Long-form Context and Query Dependent Modeling in LLMs 提出LCIRC，通过循环压缩和查询依赖建模高效处理LLM中的长文本上下文。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (9 篇)

#题目一句话要点标签🔗⭐
37 RALLRec: Improving Retrieval Augmented Large Language Model Recommendation with Representation Learning RALLRec：融合表征学习的检索增强大语言模型推荐系统 representation learning large language model ✅
38 IRepair: An Intent-Aware Approach to Repair Data-Driven Errors in Large Language Models 提出IRepair以解决大语言模型中的数据驱动错误问题 direct preference optimization large language model
39 K-ON: Stacking Knowledge On the Head Layer of Large Language Model K-ON：通过在大型语言模型的头部堆叠知识来解决KG与自然语言的粒度不匹配问题 representation learning large language model
40 Exploring the Limit of Outcome Reward for Learning Mathematical Reasoning OREAL：探索基于结果奖励的强化学习在数学推理中的极限 reinforcement learning behavior cloning distillation ✅
41 Rationalization Models for Text-to-SQL 提出基于CoT的文本到SQL生成框架，提升复杂查询的执行精度和可解释性。 distillation large language model chain-of-thought
42 Who Taught You That? Tracing Teachers in Model Distillation 提出一种基于词汇特征的教师模型溯源方法，用于识别学生模型的知识来源。 distillation instruction following
43 Optimizing Knowledge Integration in Retrieval-Augmented Generation with Self-Selection 提出Self-Selection RAG框架，通过自选择机制优化检索增强生成中的知识融合。 DPO direct preference optimization large language model
44 Ignore the KL Penalty! Boosting Exploration on Critical Tokens to Enhance RL Fine-Tuning 通过强化关键token探索，忽略KL惩罚提升RL微调效果 reinforcement learning large language model
45 C-3PO: Compact Plug-and-Play Proxy Optimization to Achieve Human-like Retrieval-Augmented Generation 提出C-3PO框架，通过轻量级代理优化实现类人检索增强生成 reinforcement learning large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页