cs.CL（2024-06-12）

📊 共 31 篇论文 | 🔗 8 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (23 🔗7) 支柱二：RL算法与架构 (RL & Architecture) (7 🔗1) 支柱五：交互与反应 (Interaction & Reaction) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (23 篇)

#题目一句话要点标签🔗⭐
1 Large Language Models Meet Text-Centric Multimodal Sentiment Analysis: A Survey 综述：大语言模型在文本中心多模态情感分析中的应用与潜力 large language model multimodal
2 Language Model Council: Democratically Benchmarking Foundation Models on Highly Subjective Tasks 提出语言模型委员会（LMC），民主化地评估大模型在主观任务上的表现 large language model foundation model
3 TasTe: Teaching Large Language Models to Translate through Self-Reflection TasTe：通过自反思教学大型语言模型进行翻译 large language model instruction following ✅
4 cPAPERS: A Dataset of Situated and Multimodal Interactive Conversations in Scientific Papers 提出cPAPERS数据集，用于科学论文中情境化多模态交互式对话研究 large language model multimodal
5 Do as I do (Safely): Mitigating Task-Specific Fine-tuning Risks in Large Language Models 提出混合安全数据微调方法，缓解大语言模型任务特定微调中的安全风险 large language model instruction following
6 Multimodal Table Understanding 提出多模态表格理解任务与Table-LLaVA模型，解决现实场景中表格图像理解难题。 large language model multimodal ✅
7 CS-Bench: A Comprehensive Benchmark for Large Language Models towards Computer Science Mastery CS-Bench：一个面向计算机科学领域的大语言模型综合评测基准 large language model ✅
8 Leveraging Large Language Models for Web Scraping 利用大型语言模型进行网页抓取，提升RAG模型在非结构化数据抽取中的效率。 large language model
9 Analyzing Large Language Models for Classroom Discussion Assessment 利用大型语言模型评估课堂讨论质量，并分析任务形式、上下文长度和少量样本的影响。 large language model
10 Adversarial Evasion Attack Efficiency against Large Language Models 研究针对大型语言模型的情感分类任务的对抗攻击效率 large language model
11 Large Language Model Unlearning via Embedding-Corrupted Prompts 提出Embedding-COrrupted Prompts以解决大语言模型知识遗忘问题 large language model ✅
12 Are Large Language Models Good Statisticians? 提出StatQA基准，评估大语言模型在统计分析和假设检验中的能力。 large language model ✅
13 MobileAIBench: Benchmarking LLMs and LMMs for On-Device Use Cases MobileAIBench：移动端LLM/LMM基准测试框架，评估量化影响与设备性能。 large language model multimodal
14 Reversing the Forget-Retain Objectives: An Efficient LLM Unlearning Framework from Logit Difference 提出ULD框架，通过Logit差分实现高效LLM知识遗忘，解决传统方法退化输出和灾难性遗忘问题。 large language model ✅
15 Next-Generation Database Interfaces: A Survey of LLM-based Text-to-SQL 综述LLM驱动的文本到SQL生成技术以应对复杂数据库查询挑战 large language model ✅
16 Is Programming by Example solved by LLMs? 评估大型语言模型在编程范例学习中的能力与局限性 large language model
17 Understanding Sounds, Missing the Questions: The Challenge of Object Hallucination in Large Audio-Language Models 揭示大型音频语言模型中的对象幻觉问题及判别性查询的挑战 large language model
18 CoXQL: A Dataset for Parsing Explanation Requests in Conversational XAI Systems 提出CoXQL数据集，用于解析会话式可解释AI系统中的解释请求。 large language model
19 AustroTox: A Dataset for Target-Based Austrian German Offensive Language Detection 提出AustroTox数据集，用于奥地利德语攻击性语言检测，并提供目标级别的标注。 large language model
20 Defining and Detecting Vulnerability in Human Evaluation Guidelines: A Preliminary Study Towards Reliable NLG Evaluation 构建评估指南漏洞数据集，提出漏洞检测方法，提升NLG评测可靠性 large language model
21 DeTriever: Decoder-representation-based Retriever for Improving NL2SQL In-Context Learning DeTriever：一种基于解码器表征的检索器，用于提升NL2SQL的上下文学习效果。 large language model
22 Watermarking Language Models with Error Correcting Codes 提出基于纠错码的语言模型水印方法，提升水印的鲁棒性和隐蔽性 large language model
23 Making Task-Oriented Dialogue Datasets More Natural by Synthetically Generating Indirect User Requests 提出基于LLM的IUR生成流程，并构建IndirectRequests数据集，提升小模型在任务型对话中处理间接用户请求的能力。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (7 篇)

#题目一句话要点标签🔗⭐
24 Mistral-C2F: Coarse to Fine Actor for Analytical and Reasoning Enhancement in RLHF and Effective-Merged LLMs 提出Mistral-C2F模型，通过粗到精Actor提升小规模LLM在RLHF中的分析推理能力。 RLHF large language model
25 It Takes Two: On the Seamlessness between Reward and Policy Model in RLHF 提出SEAM指标，衡量并提升RLHF中奖励模型与策略模型的无缝衔接 reinforcement learning RLHF
26 Guiding Frame-Level CTC Alignments Using Self-knowledge Distillation 提出一种基于自知识蒸馏的CTC对齐引导方法，提升语音识别性能。 teacher-student distillation
27 HelpSteer2: Open-source dataset for training top-performing reward models 发布HelpSteer2：用于训练高性能奖励模型的开源数据集 RLHF large language model ✅
28 Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothing Magpie：通过提示对齐的LLM，从零开始合成对齐数据，提升模型性能。 direct preference optimization large language model
29 Prompt-Based Length Controlled Generation with Multiple Control Types 提出一种基于Prompt的长度可控生成方法，解决GPT模型在多种控制类型下的长度控制问题。 reinforcement learning large language model
30 Supportiveness-based Knowledge Rewriting for Retrieval-augmented Language Modeling 提出基于支持度的知识重写方法SKR，提升检索增强语言模型生成质量。 DPO direct preference optimization

🔬 支柱五：交互与反应 (Interaction & Reaction) (1 篇)

#题目一句话要点标签🔗⭐
31 Exploring Speech Foundation Models for Speaker Diarization in Child-Adult Dyadic Interactions 利用语音基础模型提升儿童-成人交互场景下的说话人分离性能 dyadic interaction foundation model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页