cs.CL（2025-02-05）

📊 共 41 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) (34 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (6 🔗2) 支柱四：生成式动作 (Generative Motion) (1)

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

#题目一句话要点标签🔗⭐
1 On Fairness of Unified Multimodal Large Language Model for Image Generation 针对统一多模态大语言模型图像生成中的偏见问题，提出定位-修复策略与平衡偏好模型。 large language model multimodal
2 Position: Multimodal Large Language Models Can Significantly Advance Scientific Reasoning 多模态大语言模型显著提升科学推理能力，助力AGI实现 large language model multimodal
3 Enhancing Reasoning to Adapt Large Language Models for Domain-Specific Applications 提出SOLOMON架构，提升大语言模型在半导体布局设计等领域特定任务的推理适应性 large language model foundation model
4 Teaching Large Language Models Number-Focused Headline Generation With Key Element Rationales 提出基于关键要素推理链的框架，提升大语言模型在数字敏感新闻标题生成中的文本质量和数值准确性。 large language model chain-of-thought
5 Limitations of Large Language Models in Clinical Problem-Solving Arising from Inflexible Reasoning M-ARC揭示大语言模型在临床问题解决中因思维定势导致的推理局限性 large language model
6 SPRI: Aligning Large Language Models with Context-Situated Principles SPRI：通过情境化原则对齐大型语言模型，无需人工干预。 large language model ✅
7 Electronic Circuit Principles of Large Language Models 提出电子电路原理（ECP）以预测和优化大语言模型推理性能。 large language model
8 Improve Decoding Factuality by Token-wise Cross Layer Entropy of Large Language Models 提出跨层熵增强解码（END）方法，提升大语言模型生成内容的真实性。 large language model
9 Scalable In-Context Learning on Tabular Data via Retrieval-Augmented Large Language Models 提出检索增强的大语言模型，解决表格数据上上下文学习的扩展性问题 large language model
10 MedBioLM: Optimizing Medical and Biological QA with Fine-Tuned Large Language Models and Retrieval-Augmented Generation MedBioLM：通过微调LLM和RAG优化医学和生物领域的问答 large language model
11 LLM-KT: Aligning Large Language Models with Knowledge Tracing using a Plug-and-Play Instruction LLM-KT：利用即插即用指令对齐大语言模型与知识追踪 large language model
12 Sovereign Large Language Models: Advantages, Strategy and Regulations 分析大型语言模型发展趋势与战略，为国家层面AI项目实施提供政策建议 large language model
13 Context-Preserving Gradient Modulation for Large Language Models: A Novel Approach to Semantic Consistency in Long-Form Text Generation 提出上下文保持的梯度调制方法，解决长文本生成中的语义一致性问题 large language model
14 Structured Token Retention and Computational Memory Paths in Large Language Models 提出结构化Token保留与计算记忆路径，提升大语言模型长序列处理效率 large language model
15 Position: Editing Large Language Models Poses Serious Safety Risks 知识编辑技术对大语言模型构成严重安全风险 large language model
16 LLaVAC: Fine-tuning LLaVA as a Multimodal Sentiment Classifier LLaVAC：通过微调LLaVA实现多模态情感分类 multimodal ✅
17 SimMark: A Robust Sentence-Level Similarity-Based Watermarking Algorithm for Large Language Models SimMark：一种基于句子级语义相似度的鲁棒大语言模型水印算法 large language model
18 LIMO: Less is More for Reasoning LIMO：少量样本即可激发大语言模型的复杂推理能力 large language model foundation model
19 LLMs can be easily Confused by Instructional Distractions DIM-Bench基准测试揭示LLM易受指令干扰，影响任务执行 large language model instruction following
20 IAO Prompting: Making Knowledge Flow Explicit in LLMs through Structured Reasoning Templates IAO Prompting：通过结构化推理模板显式化LLM中的知识流动 large language model chain-of-thought
21 iVISPAR -- An Interactive Visual-Spatial Reasoning Benchmark for VLMs iVISPAR：一个用于评估视觉语言模型交互式视觉空间推理能力的基准 multimodal
22 Reflection-Window Decoding: Text Generation with Selective Refinement 提出反射窗口解码方法，通过选择性修正提升LLM文本生成质量 large language model
23 In Praise of Stubbornness: An Empirical Case for Cognitive-Dissonance Aware Continual Update of Knowledge in LLMs 揭示LLM持续学习中认知失调问题，提出矛盾信息检测机制 large language model
24 Diversity as a Reward: Fine-Tuning LLMs on a Mixture of Domain-Undetermined Data 提出基于多样性奖励的LLM微调方法，提升领域未定数据的性能 large language model
25 An Analysis for Reasoning Bias of Language Models with Small Initialization 研究发现小初始化规模的LLM更擅长推理任务，大初始化规模更擅长记忆任务 large language model
26 FedP$^2$EFT: Federated Learning to Personalize PEFT for Multilingual LLMs FedP$^2$EFT：联邦学习个性化PEFT多语言LLM，解决低资源语言问题 large language model
27 MEETING DELEGATE: Benchmarking LLMs on Attending Meetings on Our Behalf 提出 MEETING DELEGATE 基准测试，评估LLM在会议代理场景下的性能 large language model
28 A Benchmark for the Detection of Metalinguistic Disagreements between LLMs and Knowledge Graphs 提出用于检测LLM与知识图谱之间元语言分歧的基准测试方法 large language model
29 Speculative Prefill: Turbocharging TTFT with Lightweight and Training-Free Token Importance Estimation SpecPrefill：一种轻量级、免训练的token重要性估计方法，加速LLM的TTFT。 large language model
30 Which Words Matter Most in Zero-Shot Prompts? 提出ZIP评分以量化零样本提示中各词的重要性，揭示提示工程的内在机制。 large language model
31 ScholaWrite: A Dataset of End-to-End Scholarly Writing Process ScholaWrite：构建端到端学术写作过程数据集，助力写作助手开发 large language model
32 Lowering the Barrier of Machine Learning: Achieving Zero Manual Labeling in Review Classification Using LLMs 提出一种基于LLM的零标注评论分类方法，降低中小企业应用门槛。 large language model
33 CAMI: A Counselor Agent Supporting Motivational Interviewing through State Inference and Topic Exploration CAMI：通过状态推断和主题探索，支持动机访谈的咨询代理 large language model
34 MARAGE: Transferable Multi-Model Adversarial Attack for Retrieval-Augmented Generation Data Extraction 提出MARAGE框架，通过可迁移对抗攻击实现RAG系统的数据提取。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (6 篇)

#题目一句话要点标签🔗⭐
35 Demystifying Long Chain-of-Thought Reasoning in LLMs 探究LLM中长链思维推理的机制，揭示训练策略的关键因素。 reinforcement learning reward shaping large language model ✅
36 Advancing Reasoning in Large Language Models: Promising Methods and Approaches 综述大型语言模型推理能力提升方法，探索未来研究方向 reinforcement learning large language model chain-of-thought
37 Knowledge Distillation from Large Language Models for Household Energy Modeling 利用大型语言模型进行知识蒸馏，用于家庭能源建模，生成多样化数据。 distillation large language model ✅
38 DreamDPO: Aligning Text-to-3D Generation with Human Preferences via Direct Preference Optimization DreamDPO：通过直接偏好优化对齐文本到3D生成与人类偏好 direct preference optimization multimodal
39 Training an LLM-as-a-Judge Model: Pipeline, Insights, and Practical Lessons 提出Themis：一种可进行复杂情境感知评估的LLM评判模型 distillation large language model instruction following
40 Out-of-Distribution Detection using Synthetic Data Generation 利用LLM生成合成数据，提升文本分类系统中的OOD检测性能 RLHF large language model

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
41 Token Assorted: Mixing Latent and Text Tokens for Improved Language Model Reasoning 提出Token Assorted方法，通过混合隐变量和文本token提升语言模型推理能力。 VQ-VAE large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页