cs.CL（2024-07-18）

📊 共 27 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (22 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (3) 支柱一：机器人控制 (Robot Control) (1) 支柱四：生成式动作 (Generative Motion) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (22 篇)

#题目一句话要点标签🔗⭐
1 Crafting Efficient Fine-Tuning Strategies for Large Language Models 提出高效微调策略，降低大语言模型微调的数据需求和计算成本 large language model
2 Evaluating Large Language Models for Anxiety and Depression Classification using Counseling and Psychotherapy Transcripts 评估大型语言模型在咨询记录中焦虑和抑郁分类的有效性 large language model
3 An Application of Large Language Models to Coding Negotiation Transcripts 利用大型语言模型分析谈判记录，探索LLM在现实场景中的应用潜力。 large language model
4 Combining Constraint Programming Reasoning with Large Language Model Predictions 结合约束编程与大语言模型预测，提升约束条件下文本生成质量与效率 large language model
5 End-To-End Clinical Trial Matching with Large Language Models 利用大型语言模型实现端到端临床试验匹配，提升癌症患者治疗效率。 large language model
6 Translate-and-Revise: Boosting Large Language Models for Constrained Translation 提出Translate-and-Revise框架，提升大语言模型在约束翻译任务中的性能。 large language model
7 Are Large Language Models Capable of Generating Human-Level Narratives? 分析LLM生成叙事能力，揭示其在故事发展和情感表达上的局限性 large language model
8 PM-LLM-Benchmark: Evaluating Large Language Models on Process Mining Tasks PM-LLM-Benchmark：评估大型语言模型在流程挖掘任务中的性能 large language model
9 Dynamic Sentiment Analysis with Local Large Language Models using Majority Voting: A Study on Factors Affecting Restaurant Evaluation 提出基于多数投票的局部大语言模型动态情感分析方法，提升餐厅评价的鲁棒性。 large language model
10 Prover-Verifier Games improve legibility of LLM outputs 提出基于证明-验证者游戏的训练算法以提高LLM输出的可读性 large language model chain-of-thought
11 Learning-From-Mistakes Prompting for Indigenous Language Translation 提出Learning-From-Mistakes Prompting方法，提升极低资源土著语言翻译质量。 large language model chain-of-thought
12 From Words to Worlds: Compositionality for Cognitive Architectures 研究LLM的组合性学习能力，揭示规模化与指令调优的微妙影响。 large language model
13 Werewolf Arena: A Case Study in LLM Evaluation via Social Deduction 提出狼人竞技场：通过社交推理游戏评估大型语言模型 large language model
14 Weak-to-Strong Reasoning 提出一种渐进式学习框架，提升大语言模型在复杂推理任务中的能力。 large language model ✅
15 Scaling Laws with Vocabulary: Larger Models Deserve Larger Vocabularies 提出基于计算预算优化词汇表大小的方法，提升大语言模型预训练效率。 large language model ✅
16 How Reliable are LLMs as Knowledge Bases? Re-thinking Facutality and Consistency 重新评估LLM作为知识库的可靠性，关注事实性和一致性 large language model
17 FANTAstic SEquences and Where to Find Them: Faithful and Efficient API Call Generation through State-tracked Constrained Decoding and Reranking 提出FANTASE框架，通过状态追踪约束解码和重排序，实现API调用的忠实高效生成。 large language model
18 Baba Is AI: Break the Rules to Beat the Benchmark 提出Baba Is You游戏新基准，揭示LLM在规则操纵与泛化上的局限性 large language model
19 Research on Tibetan Tourism Viewpoints information generation system based on LLM 提出DualGen Bridge AI系统，提升LLM在藏区旅游信息生成中的性能 large language model
20 Attention Overflow: Language Model Input Blur during Long-Context Missing Items Recommendation 揭示长文本输入下语言模型“注意力溢出”问题，影响推荐性能 large language model
21 SpeciaLex: A Benchmark for In-Context Specialized Lexicon Learning SpeciaLex：一个用于评估LLM在上下文学习中专业词汇理解能力的基准 large language model
22 Retrieval-Augmented Generation for Natural Language Processing: A Survey 综述检索增强生成（RAG）技术，解决大语言模型幻觉、知识更新和领域知识不足问题。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (3 篇)

#题目一句话要点标签🔗⭐
23 Understanding Reference Policies in Direct Preference Optimization 研究DPO中参考策略的影响，揭示其对性能的制约与优化策略 DPO direct preference optimization large language model
24 BiasDPO: Mitigating Bias in Language Models through Direct Preference Optimization BiasDPO：通过直接偏好优化缓解语言模型中的偏见 DPO direct preference optimization large language model
25 Learning Goal-Conditioned Representations for Language Reward Models 提出目标条件对比学习方法，提升语言模型奖励模型的性能和可控性。 reinforcement learning RLHF representation learning

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
26 Black-Box Opinion Manipulation Attacks to Retrieval-Augmented Generation of Large Language Models 针对RAG模型的黑盒观点操纵攻击，揭示其认知偏见风险 manipulation large language model

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
27 Towards Zero-Shot Multimodal Machine Translation 提出ZeroMMT，利用多模态英语数据实现零样本多模态机器翻译。 classifier-free guidance multimodal

⬅️ 返回 cs.CL 首页 · 🏠 返回主页