cs.CL（2024-05-08）

📊 共 25 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)

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

#题目一句话要点标签🔗⭐
1 Empathy Through Multimodality in Conversational Interfaces 提出基于LLM的多模态对话健康助手，提升心理健康支持中的共情能力 large language model multimodal
2 Red-Teaming for Inducing Societal Bias in Large Language Models 提出情感偏见探测与偏见知识图谱方法，诱导大语言模型产生社会偏见。 large language model
3 Utilizing Large Language Models to Generate Synthetic Data to Increase the Performance of BERT-Based Neural Networks 利用大型语言模型生成合成数据提升BERT模型在医疗领域的性能 large language model
4 Mitigating Exaggerated Safety in Large Language Models 提出多策略Prompting方法，有效缓解大语言模型过度安全问题 large language model
5 Fishing for Magikarp: Automatically Detecting Under-trained Tokens in Large Language Models 提出自动检测方法，识别大型语言模型中欠训练的token，提升模型安全性和效率。 large language model
6 P-ICL: Point In-Context Learning for Named Entity Recognition with Large Language Models 提出P-ICL框架，利用关键实体信息增强大语言模型在命名实体识别中的上下文学习能力。 large language model
7 Zero-shot LLM-guided Counterfactual Generation: A Case Study on NLP Model Evaluation 提出基于零样本LLM引导的反事实生成方法，用于NLP模型评估。 large language model instruction following
8 Poser: Unmasking Alignment Faking LLMs by Manipulating Their Internals Poser：通过操纵LLM内部机制揭示伪装对齐行为 large language model
9 Evaluating Students' Open-ended Written Responses with LLMs: Using the RAG Framework for GPT-3.5, GPT-4, Claude-3, and Mistral-Large 利用RAG框架评估LLM在开放式学生答案评估中的表现 large language model
10 "They are uncultured": Unveiling Covert Harms and Social Threats in LLM Generated Conversations 提出CHAST指标体系，揭示LLM生成对话中针对非西方文化概念的隐蔽偏见与社会威胁 large language model
11 The Effect of Model Size on LLM Post-hoc Explainability via LIME 研究表明增大LLM模型尺寸不一定提升LIME事后解释的合理性 large language model
12 Open Source Language Models Can Provide Feedback: Evaluating LLMs' Ability to Help Students Using GPT-4-As-A-Judge 利用GPT-4评估开源LLM在编程教育反馈中的表现，探索其负责任应用 large language model
13 You Only Cache Once: Decoder-Decoder Architectures for Language Models YOCO：一种仅缓存一次键值对的Decoder-Decoder架构，提升大语言模型效率。 large language model
14 LLMs with Personalities in Multi-issue Negotiation Games 利用人格化LLM进行多议题谈判博弈，提升谈判策略设计 large language model
15 MIDGARD: Self-Consistency Using Minimum Description Length for Structured Commonsense Reasoning 提出MIDGARD，利用最小描述长度进行自洽性推理，解决结构化常识推理中的误差传播问题。 large language model
16 AirGapAgent: Protecting Privacy-Conscious Conversational Agents AirGapAgent：保护隐私敏感型对话Agent，防止上下文劫持 large language model
17 XAMPLER: Learning to Retrieve Cross-Lingual In-Context Examples XAMPLER：学习检索跨语言上下文示例，提升低资源语言的上下文学习能力 large language model ✅
18 QFMTS: Generating Query-Focused Summaries over Multi-Table Inputs QFMTS：提出一种基于查询的多表格输入摘要生成方法，提升信息需求满足度。 large language model
19 Seeds of Stereotypes: A Large-Scale Textual Analysis of Race and Gender Associations with Diseases in Online Sources 大规模文本分析揭示在线资源中疾病与种族、性别刻板印象的关联 large language model
20 APrompt4EM: Augmented Prompt Tuning for Generalized Entity Matching 提出APrompt4EM，通过增强Prompt Tuning解决广义实体匹配中的低资源问题。 large language model
21 ACORN: Aspect-wise Commonsense Reasoning Explanation Evaluation 提出ACORN数据集，用于评估LLM在常识推理解释质量评估中的表现。 large language model ✅
22 Automated Conversion of Static to Dynamic Scheduler via Natural Language 提出RAGDyS框架，利用自然语言将静态调度器自动转换为动态调度器 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 Conversational Topic Recommendation in Counseling and Psychotherapy with Decision Transformer and Large Language Models 利用决策Transformer和LLM，为心理咨询对话提供话题推荐 reinforcement learning offline reinforcement learning decision transformer
24 CourseGPT-zh: an Educational Large Language Model Based on Knowledge Distillation Incorporating Prompt Optimization 提出CourseGPT-zh，一种基于知识蒸馏和提示优化的教育领域大语言模型。 distillation large language model
25 ADELIE: Aligning Large Language Models on Information Extraction ADELIE：通过对齐大型语言模型进行信息抽取，实现SOTA性能。 DPO direct preference optimization large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页