cs.CL（2025-02-04）

📊 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) (40 🔗8) 支柱二：RL算法与架构 (RL & Architecture) (5 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (40 篇)

#题目一句话要点标签🔗⭐
1 SAISA: Towards Multimodal Large Language Models with Both Training and Inference Efficiency 提出SAISA，一种提升训练和推理效率的多模态大语言模型架构 large language model multimodal ✅
2 CognArtive: Large Language Models for Automating Art Analysis and Decoding Aesthetic Elements CognArtive：利用大型语言模型自动化艺术分析与解码审美元素 large language model multimodal
3 Open Foundation Models in Healthcare: Challenges, Paradoxes, and Opportunities with GenAI Driven Personalized Prescription 探索开源大模型在医疗领域的潜力：以个性化处方为例 large language model foundation model
4 LLM-ProS: Analyzing Large Language Models' Performance in Competitive Problem Solving LLM-ProS：评估大语言模型在竞争性问题求解中的性能 large language model chain-of-thought
5 Boosting Multimodal Reasoning with Automated Structured Thinking 提出AStar：通过自动化结构化思考提升多模态推理能力 large language model multimodal
6 JingFang: An Expert-Level Large Language Model for Traditional Chinese Medicine Clinical Consultation and Syndrome Differentiation-Based Treatment JingFang：一种专家级中医临床诊疗与辨证论治大语言模型 large language model chain-of-thought
7 Investigating the Robustness of Deductive Reasoning with Large Language Models 研究大型语言模型在演绎推理任务中的鲁棒性，并分析不同方法组件的影响。 large language model
8 Multilingual Machine Translation with Open Large Language Models at Practical Scale: An Empirical Study 提出PFMS数据混合策略，GemmaX2-28在多语言机器翻译上达到领先水平。 large language model
9 Generative Psycho-Lexical Approach for Constructing Value Systems in Large Language Models 提出生成式心理词汇方法，构建更符合LLM特性的价值体系 large language model
10 Activation-Informed Merging of Large Language Models 提出激活信息引导的模型合并方法AIM，提升大语言模型性能和鲁棒性 large language model
11 Prompt-based Depth Pruning of Large Language Models 提出PuDDing：一种基于Prompt路由的大语言模型动态深度剪枝方法 large language model
12 CoAT: Chain-of-Associated-Thoughts Framework for Enhancing Large Language Models Reasoning 提出CoAT框架，结合MCTS与联想记忆增强大语言模型推理能力 large language model
13 Evalita-LLM: Benchmarking Large Language Models on Italian Evalita-LLM：意大利语大型语言模型评测基准，解决文化偏见和提示敏感性问题。 large language model
14 Robust and Secure Code Watermarking for Large Language Models via ML/Crypto Codesign RoSeMary：面向大语言模型代码水印的ML/密码学协同设计框架 large language model
15 Automating Mathematical Proof Generation Using Large Language Model Agents and Knowledge Graphs KG-Prover：利用知识图谱增强大语言模型，实现自动化数学证明生成 large language model
16 Contextual Memory Reweaving in Large Language Models Using Layered Latent State Reconstruction 提出上下文记忆重织框架，通过分层潜在状态重构增强大语言模型长程记忆能力 large language model
17 Gradient-Regularized Latent Space Modulation in Large Language Models for Structured Contextual Synthesis 提出梯度正则化潜在空间调制方法，用于大语言模型中生成结构化文本。 large language model
18 CITER: Collaborative Inference for Efficient Large Language Model Decoding with Token-Level Routing CITER：通过令牌级路由实现高效大语言模型协同推理 large language model ✅
19 Conceptual Metaphor Theory as a Prompting Paradigm for Large Language Models 提出基于概念隐喻理论的认知提示方法，提升大语言模型在复杂推理任务中的表现。 large language model
20 SmolLM2: When Smol Goes Big -- Data-Centric Training of a Small Language Model SmolLM2：通过数据为中心训练，打造高性能小规模语言模型 large language model instruction following
21 STAIR: Improving Safety Alignment with Introspective Reasoning STAIR：通过内省推理提升大语言模型的安全对齐 large language model chain-of-thought ✅
22 Premise-Augmented Reasoning Chains Improve Error Identification in Math reasoning with LLMs 提出PARC：通过前提增强推理链提升LLM数学推理中的错误识别 large language model chain-of-thought
23 Position: Scaling LLM Agents Requires Asymptotic Analysis with LLM Primitives 利用LLM原语的渐近分析，扩展LLM Agent规模 large language model
24 Transformers Boost the Performance of Decision Trees on Tabular Data across Sample Sizes 提出LLM-Boost和PFN-Boost，融合Transformer与GBDT提升表格数据在不同样本量下的性能。 large language model ✅
25 A Training-Free Length Extrapolation Approach for LLMs: Greedy Attention Logit Interpolation (GALI) 提出GALI，一种免训练的长文本LLM长度外推方法，解决注意力机制失效问题。 large language model ✅
26 Are Language Models Up to Sequential Optimization Problems? From Evaluation to a Hegelian-Inspired Enhancement 提出WorldGen与ACE框架以提升语言模型在序列优化问题中的表现 large language model
27 Adaptive Self-improvement LLM Agentic System for ML Library Development 提出自适应自提升LLM Agent系统，用于机器学习库的自动开发 large language model
28 Dynamic benchmarking framework for LLM-based conversational data capture 提出一种动态基准测试框架，用于评估基于LLM的对话数据捕获能力。 large language model
29 SCALM: Detecting Bad Practices in Smart Contracts Through LLMs SCALM：利用大型语言模型检测智能合约中的不良实践 large language model
30 ReSpark: Leveraging Previous Data Reports as References to Generate New Reports with LLMs ReSpark：利用LLM和历史报告生成新数据报告，降低分析门槛 large language model
31 Mass-Editing Memory with Attention in Transformers: A cross-lingual exploration of knowledge 提出MEMAT方法，利用Transformer注意力机制实现大规模、跨语言的知识编辑。 large language model ✅
32 Rethinking stance detection: A theoretically-informed research agenda for user-level inference using language models 重新思考立场检测：一个基于语言模型的用户级别推断的理论驱动研究议程 large language model
33 ASCenD-BDS: Adaptable, Stochastic and Context-aware framework for Detection of Bias, Discrimination and Stereotyping 提出ASCenD-BDS框架，用于自适应、随机和上下文感知地检测LLM中的偏见、歧视和刻板印象。 large language model
34 TRUTH DECAY: Quantifying Multi-Turn Sycophancy in Language Models 提出TRUTH DECAY基准，量化语言模型在多轮对话中的谄媚衰减现象。 large language model
35 M2R2: Mixture of Multi-Rate Residuals for Efficient Transformer Inference M2R2：提出混合多速率残差框架，提升Transformer推理效率。 large language model
36 Reasoning Bias of Next Token Prediction Training 揭示NTP训练的推理偏好：噪声正则化提升LLM泛化与鲁棒性 large language model
37 Can LLMs Assist Annotators in Identifying Morality Frames? -- Case Study on Vaccination Debate on Social Media 利用大型语言模型辅助标注员识别社交媒体疫苗辩论中的道德框架 large language model
38 Multi-Lingual Cyber Threat Detection in Tweets/X Using ML, DL, and LLM: A Comparative Analysis 提出基于Bi-LSTM的多语言网络威胁检测方法，提升社交媒体内容安全。 large language model ✅
39 Token Cleaning: Fine-Grained Data Selection for LLM Supervised Fine-Tuning 提出Token Cleaning方法，通过细粒度数据选择提升LLM监督微调效果 large language model ✅
40 Can LLMs Maintain Fundamental Abilities under KV Cache Compression? 针对KV缓存压缩对LLM能力的影响，提出ShotKV压缩方法，提升长文本生成性能。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (5 篇)

#题目一句话要点标签🔗⭐
41 Reusing Embeddings: Reproducible Reward Model Research in Large Language Model Alignment without GPUs 提出基于嵌入的奖励模型，加速LLM对齐研究并提高可复现性 reinforcement learning RLHF large language model
42 Reviving The Classics: Active Reward Modeling in Large Language Model Alignment 提出基于Fisher信息的奖励模型主动学习方法，提升大语言模型对齐效率 reinforcement learning RLHF large language model
43 Satori: Reinforcement Learning with Chain-of-Action-Thought Enhances LLM Reasoning via Autoregressive Search Satori：通过行动-思考链增强LLM推理能力的自回归搜索强化学习 reinforcement learning large language model
44 CodeSteer: Symbolic-Augmented Language Models via Code/Text Guidance CodeSteer通过代码/文本引导增强LLM的符号计算能力，显著提升复杂任务性能。 DPO direct preference optimization large language model ✅
45 LongDPO: Unlock Better Long-form Generation Abilities for LLMs via Critique-augmented Stepwise Information LongDPO：通过评论增强的逐步信息，提升LLM的长文本生成能力 preference learning DPO

⬅️ 返回 cs.CL 首页 · 🏠 返回主页