cs.CL（2024-10-28）

📊 共 43 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) (31 🔗6) 支柱二：RL算法与架构 (RL & Architecture) (11 🔗3) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (31 篇)

#题目一句话要点标签🔗⭐
1 Large Language Model Benchmarks in Medical Tasks 综述医学领域大语言模型评测基准，促进临床任务的LLM应用。 large language model multimodal
2 TransformLLM: Adapting Large Language Models via LLM-Transformed Reading Comprehension Text 提出TransformLLM，通过LLM转换的阅读理解文本来适配大型语言模型，提升其在特定领域的性能。 large language model
3 Simple Is Effective: The Roles of Graphs and Large Language Models in Knowledge-Graph-Based Retrieval-Augmented Generation SubgraphRAG：利用图结构和轻量级模型提升知识图谱检索增强生成效果 large language model
4 Thank You, Stingray: Multilingual Large Language Models Can Not (Yet) Disambiguate Cross-Lingual Word Sense StingrayBench：揭示多语言大模型在跨语言词义消歧方面的局限性 large language model
5 Group-SAE: Efficient Training of Sparse Autoencoders for Large Language Models via Layer Groups 提出Group-SAE以解决大语言模型稀疏自编码器训练效率问题 large language model
6 Can Large Language Models Act as Symbolic Reasoners? 研究大型语言模型是否具备符号推理能力及其可解释性 large language model
7 CT2C-QA: Multimodal Question Answering over Chinese Text, Table and Chart 提出C$ ext{T}^2$C-QA数据集与AED多智能体系统，用于解决中文文本、表格和图表的多模态问答问题。 multimodal
8 LLMCBench: Benchmarking Large Language Model Compression for Efficient Deployment LLMCBench：构建大语言模型压缩基准，促进高效部署 large language model ✅
9 A Survey on Automatic Credibility Assessment Using Textual Credibility Signals in the Era of Large Language Models 在大语言模型时代，综述基于文本可信度信号的自动可信度评估方法。 large language model
10 An Actor-Critic Approach to Boosting Text-to-SQL Large Language Model 提出Actor-Critic框架，提升大语言模型在Text-to-SQL任务中的性能 large language model
11 CRAT: A Multi-Agent Framework for Causality-Enhanced Reflective and Retrieval-Augmented Translation with Large Language Models 提出CRAT多Agent框架，增强LLM在机器翻译中对上下文相关术语的处理能力 large language model
12 NewTerm: Benchmarking Real-Time New Terms for Large Language Models with Annual Updates NewTerm：构建年度更新的LLM实时新词评测基准，解决知识截断问题。 large language model ✅
13 Rephrasing natural text data with different languages and quality levels for Large Language Model pre-training 提出多语言和多质量等级的文本复述方法，用于提升大型语言模型预训练效果 large language model
14 ElectionSim: Massive Population Election Simulation Powered by Large Language Model Driven Agents 提出ElectionSim，基于大语言模型驱动的Agent进行大规模选举模拟。 large language model
15 SandboxAQ's submission to MRL 2024 Shared Task on Multi-lingual Multi-task Information Retrieval SandboxAQ探索多语言多任务信息检索，着重分析大语言模型在QA和NER任务上的性能差异。 large language model chain-of-thought
16 MultiTok: Variable-Length Tokenization for Efficient LLMs Adapted from LZW Compression MultiTok：一种基于LZW压缩的高效变长分词方法，加速LLM训练。 large language model
17 Can Machines Think Like Humans? A Behavioral Evaluation of LLM Agents in Dictator Games 评估大型语言模型在独裁者游戏中的亲社会行为 large language model
18 SCULPT: Systematic Tuning of Long Prompts SCULPT：通过系统调优长提示来提升大语言模型性能 large language model
19 Graph-based Uncertainty Metrics for Long-form Language Model Outputs 提出基于图的LLM不确定性度量方法，提升长文本生成的事实性和信息量。 large language model
20 Estimating Causal Effects of Text Interventions Leveraging LLMs CausalDANN：利用LLM进行文本干预因果效应估计，解决高维文本数据挑战。 large language model
21 EoRA: Fine-tuning-free Compensation for Compressed LLM with Eigenspace Low-Rank Approximation EoRA：一种免微调的低秩本征空间近似方法，用于补偿压缩LLM的精度损失。 large language model ✅
22 Palisade -- Prompt Injection Detection Framework Palisade：一种用于检测提示注入攻击的多层防御框架 large language model
23 FACT: Examining the Effectiveness of Iterative Context Rewriting for Multi-fact Retrieval 提出FACT迭代上下文重写方法，解决LLM多事实检索中“中间信息丢失”问题 large language model
24 FATH: Authentication-based Test-time Defense against Indirect Prompt Injection Attacks 提出基于哈希认证标签的FATH方法，防御针对LLM应用的间接提示注入攻击。 large language model ✅
25 Arithmetic Without Algorithms: Language Models Solve Math With a Bag of Heuristics 揭示大语言模型算术能力：并非算法或记忆，而是启发式规则组合 large language model
26 M2rc-Eval: Massively Multilingual Repository-level Code Completion Evaluation 提出M2RC-EVAL：大规模多语言仓库级代码补全评估基准 large language model
27 Autoformalize Mathematical Statements by Symbolic Equivalence and Semantic Consistency 提出基于符号等价和语义一致性的数学语句自动形式化框架 large language model
28 AutoRAG: Automated Framework for optimization of Retrieval Augmented Generation Pipeline 提出AutoRAG框架，自动优化检索增强生成（RAG）流水线，提升特定数据集性能。 large language model ✅
29 A Simple Yet Effective Corpus Construction Framework for Indonesian Grammatical Error Correction 提出印尼语语法纠错语料库构建框架，并探索LLM辅助标注可行性 large language model ✅
30 LLMs are Biased Evaluators But Not Biased for Retrieval Augmented Generation 研究表明，大语言模型在检索增强生成中作为评估者时，偏见不明显，更注重事实准确性。 large language model
31 Are LLM-Judges Robust to Expressions of Uncertainty? Investigating the effect of Epistemic Markers on LLM-based Evaluation 发现LLM评判者对不确定性表达不鲁棒：存在对认知标记的负偏见 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (11 篇)

#题目一句话要点标签🔗⭐
32 LongReward: Improving Long-context Large Language Models with AI Feedback 提出LongReward，利用AI反馈提升长文本大语言模型性能 reinforcement learning offline RL DPO
33 UFT: Unifying Fine-Tuning of SFT and RLHF/DPO/UNA through a Generalized Implicit Reward Function UFT：通过广义隐式奖励函数统一SFT与RLHF/DPO/UNA的微调 RLHF DPO instruction following
34 Stealthy Jailbreak Attacks on Large Language Models via Benign Data Mirroring 提出基于良性数据镜像的隐蔽越狱攻击方法，提升大语言模型安全性评估的隐蔽性。 distillation large language model
35 DeTeCtive: Detecting AI-generated Text via Multi-Level Contrastive Learning 提出DeTeCtive以解决AI生成文本检测的局限性问题 contrastive learning large language model ✅
36 KD-LoRA: A Hybrid Approach to Efficient Fine-Tuning with LoRA and Knowledge Distillation KD-LoRA：结合LoRA与知识蒸馏的高效微调方法，降低资源需求。 distillation large language model ✅
37 CycleResearcher: Improving Automated Research via Automated Review CycleResearcher：通过自动化评审改进自动化研究 reinforcement learning MAE large language model ✅
38 CARMO: Dynamic Criteria Generation for Context-Aware Reward Modelling CARMO：通过动态生成上下文相关标准，提升奖励模型性能并缓解奖励攻击。 reinforcement learning RLHF large language model
39 Reward Modeling with Weak Supervision for Language Models 提出基于弱监督的奖励模型训练方法，提升语言模型在人机反馈强化学习中的性能。 reinforcement learning RLHF large language model
40 Reducing the Scope of Language Models 提出语言模型范围限定方法，使其仅响应特定任务，提升部署效率。 preference learning large language model
41 Relaxed Recursive Transformers: Effective Parameter Sharing with Layer-wise LoRA 提出Relaxed Recursive Transformers，通过层间LoRA实现高效参数共享，缩小LLM体积。 distillation large language model
42 Relation-based Counterfactual Data Augmentation and Contrastive Learning for Robustifying Natural Language Inference Models 提出基于关系的对抗数据增强和对比学习，增强自然语言推理模型的鲁棒性 contrastive learning

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
43 Fine-tuned Large Language Models (LLMs): Improved Prompt Injection Attacks Detection 微调大型语言模型提升提示注入攻击检测能力 manipulation large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页