cs.CL（2024-08-05）

📊 共 26 篇论文 | 🔗 5 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (20 🔗5) 支柱二：RL算法与架构 (RL & Architecture) (6)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (20 篇)

#题目一句话要点标签🔗⭐
1 Caution for the Environment: Multimodal LLM Agents are Susceptible to Environmental Distractions 揭示多模态LLM智能体在GUI环境中易受环境干扰的问题 generalist agent large language model multimodal
2 Dialogue Ontology Relation Extraction via Constrained Chain-of-Thought Decoding 提出基于约束链式思考解码的对话本体关系抽取方法，提升泛化能力。 large language model chain-of-thought
3 Large Model Strategic Thinking, Small Model Efficiency: Transferring Theory of Mind in Large Language Models 利用大模型思维，提升小模型效率：迁移大语言模型的心理理论 large language model
4 XMainframe: A Large Language Model for Mainframe Modernization XMainframe：用于大型机现代化的专用大型语言模型 large language model
5 SEAS: Self-Evolving Adversarial Safety Optimization for Large Language Models 提出SEAS框架，通过自进化对抗安全优化提升大语言模型安全性 large language model ✅
6 A Few-Shot Approach for Relation Extraction Domain Adaptation using Large Language Models 提出一种基于大语言模型的小样本关系抽取领域自适应方法，用于科学知识图谱构建。 large language model
7 UnifiedMLLM: Enabling Unified Representation for Multi-modal Multi-tasks With Large Language Model 提出UnifiedMLLM以解决多模态任务统一表示问题 large language model ✅
8 Let Me Speak Freely? A Study on the Impact of Format Restrictions on Performance of Large Language Models 研究表明：格式限制显著降低大语言模型在推理任务中的性能 large language model
9 Why Are My Prompts Leaked? Unraveling Prompt Extraction Threats in Customized Large Language Models 揭示定制大语言模型中的提示词泄露风险并提出防御策略 large language model ✅
10 Pula: Training Large Language Models for Setswana Pula：训练用于塞茨瓦纳语的大型语言模型，性能超越GPT-4o和Gemini 1.5 Pro large language model
11 Do Large Language Models Speak All Languages Equally? A Comparative Study in Low-Resource Settings 对比研究LLM在低资源语言环境下的表现，揭示其性能差异 large language model
12 RAG Foundry: A Framework for Enhancing LLMs for Retrieval Augmented Generation RAG Foundry：用于增强LLM的检索增强生成开源框架 large language model ✅
13 Winning Amazon KDD Cup'24 针对在线购物场景，提出基于Qwen2-72B微调和数据增强的LLM智能助手方案，赢得Amazon KDD Cup'24全部任务冠军。 large language model
14 CodeACT: Code Adaptive Compute-efficient Tuning Framework for Code LLMs CodeACT：面向代码大模型的代码自适应计算高效微调框架 large language model
15 Leveraging the Power of LLMs: A Fine-Tuning Approach for High-Quality Aspect-Based Summarization 通过微调LLM，本文提出了一种高质量的基于方面的情感摘要生成方法。 large language model
16 Long Input Benchmark for Russian Analysis LIBRA：面向俄语分析的长文本输入基准评测，促进长文本理解能力评估 large language model
17 MaterioMiner -- An ontology-based text mining dataset for extraction of process-structure-property entities 提出MaterioMiner数据集，用于材料科学领域过程-结构-性质实体抽取。 large language model
18 LLM economicus? Mapping the Behavioral Biases of LLMs via Utility Theory 利用效用理论评估大语言模型的行为偏差，揭示其经济决策非完全理性或类人 large language model
19 The Mechanics of Conceptual Interpretation in GPT Models: Interpretative Insights 提出概念编辑方法，揭示GPT模型中概念理解的机制 large language model
20 ReDel: A Toolkit for LLM-Powered Recursive Multi-Agent Systems ReDel：一个支持LLM驱动的递归多智能体系统工具包，用于灵活的任务委派和组织。 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🔗⭐
21 Can Reinforcement Learning Unlock the Hidden Dangers in Aligned Large Language Models? 提出强化学习优化对抗触发器以解决大语言模型的安全问题 reinforcement learning large language model
22 Progressively Label Enhancement for Large Language Model Alignment 提出PLE框架，通过动态调整训练过程提升大语言模型对齐效果 reinforcement learning RLHF large language model
23 SNFinLLM: Systematic and Nuanced Financial Domain Adaptation of Chinese Large Language Models 提出SNFinLLM，针对中文金融领域进行系统且细致的领域自适应，提升金融计算和机器阅读理解能力。 DPO direct preference optimization large language model
24 Strong and weak alignment of large language models with human values 区分强弱价值对齐，揭示大语言模型在理解人类价值观方面的局限性 reinforcement learning large language model
25 A Framework for Fine-Tuning LLMs using Heterogeneous Feedback 提出一种利用异构反馈微调大型语言模型的框架，提升指令遵循和减少偏差。 reinforcement learning RLHF large language model
26 Evaluating and Enhancing LLMs Agent based on Theory of Mind in Guandan: A Multi-Player Cooperative Game under Imperfect Information 提出基于心智理论的LLM智能体，提升在非完美信息合作游戏掼蛋中的表现 reinforcement learning large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页