cs.CL（2024-11-07）

📊 共 27 篇论文 | 🔗 3 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (23 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (4 🔗1)

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

#题目一句话要点标签🔗⭐
1 Mixture-of-Transformers: A Sparse and Scalable Architecture for Multi-Modal Foundation Models 提出混合Transformer（MoT）架构，用于高效可扩展的多模态基础模型训练。 large language model foundation model
2 Toward Cultural Interpretability: A Linguistic Anthropological Framework for Describing and Evaluating Large Language Models (LLMs) 提出文化可解释性框架，提升LLM在文化和语言理解上的价值对齐。 large language model
3 Meta-Reasoning Improves Tool Use in Large Language Models TECTON：通过元推理提升大型语言模型工具使用能力 large language model
4 Deploying Large Language Models With Retrieval Augmented Generation 利用检索增强生成部署大型语言模型，提升信息检索的准确性和可靠性 large language model
5 OpenCoder: The Open Cookbook for Top-Tier Code Large Language Models OpenCoder：开源顶级代码大语言模型，提供可复现的训练流程与数据。 large language model
6 Prompt-Guided Internal States for Hallucination Detection of Large Language Models 提出PRISM框架，利用Prompt引导LLM内部状态，提升幻觉检测跨域泛化能力 large language model
7 Self-Calibrated Listwise Reranking with Large Language Models 提出自校准列表重排序方法以解决LLM上下文窗口限制问题 large language model
8 Best Practices for Distilling Large Language Models into BERT for Web Search Ranking 提出蒸馏技术将大型语言模型转化为BERT以优化网页搜索排名 large language model
9 Thanos: Enhancing Conversational Agents with Skill-of-Mind-Infused Large Language Model Thanos：通过注入心智技能的大语言模型增强对话智能体 large language model
10 Measuring short-form factuality in large language models 提出SimpleQA基准，用于评估大语言模型在短文本问答中的事实性能力。 large language model ✅
11 Leveraging LLMs to Enable Natural Language Search on Go-to-market Platforms 利用LLM在GTM平台上实现自然语言搜索，提升企业信息检索效率 large language model chain-of-thought
12 FMEA Builder: Expert Guided Text Generation for Equipment Maintenance FMEA Builder：专家指导下的设备维护文本生成系统 large language model foundation model
13 Bayesian Calibration of Win Rate Estimation with LLM Evaluators 提出贝叶斯校准方法，提升LLM评估器胜率估计的准确性 large language model instruction following
14 SuffixDecoding: Extreme Speculative Decoding for Emerging AI Applications 提出SuffixDecoding，利用后缀树缓存加速LLM Agent重复性推理任务。 large language model ✅
15 BitNet a4.8: 4-bit Activations for 1-bit LLMs BitNet a4.8：为1-bit LLM引入4-bit激活，提升推理效率 large language model
16 VTechAGP: An Academic-to-General-Audience Text Paraphrase Dataset and Benchmark Models 提出VTechAGP学术-通俗文本释义数据集与DSPT5模型，解决学术文本通俗化难题。 large language model
17 Explaining Mixtures of Sources in News Articles 提出新闻文章中来源选择的解释框架，通过预测来源选择模式理解记者写作计划。 large language model
18 Enabling LLM Knowledge Analysis via Extensive Materialization 通过大规模物化实现LLM知识分析，构建GPTKB知识库。 large language model
19 STAND-Guard: A Small Task-Adaptive Content Moderation Model 提出STAND-GUARD，一种小型任务自适应内容审核模型，适用于各类内容审核场景。 large language model
20 CodeLutra: Boosting LLM Code Generation via Preference-Guided Refinement CodeLutra：通过偏好引导的精炼提升LLM代码生成能力 large language model
21 Needle Threading: Can LLMs Follow Threads through Near-Million-Scale Haystacks? 评估LLM在近百万规模文档中追踪信息线程的能力，揭示有效上下文长度限制。 large language model
22 LuxBank: The First Universal Dependency Treebank for Luxembourgish 构建首个卢森堡语通用依存句法树库LuxBank，填补低资源语言句法分析空白。 large language model
23 Gradient Localization Improves Lifelong Pretraining of Language Models 提出梯度定位方法，提升语言模型终身预训练效果 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (4 篇)

#题目一句话要点标签🔗⭐
24 Performance-Guided LLM Knowledge Distillation for Efficient Text Classification at Scale 提出性能引导的LLM知识蒸馏PGKD，用于高效大规模文本分类 teacher-student distillation large language model
25 Abstract2Appendix: Academic Reviews Enhance LLM Long-Context Capabilities 利用学术评审数据增强LLM长文本处理能力，DPO方法优于SFT。 DPO direct preference optimization large language model
26 ACCIO: Table Understanding Enhanced via Contrastive Learning with Aggregations ACCIO：利用对比学习与聚合增强表格理解 contrastive learning ✅
27 One fish, two fish, but not the whole sea: Alignment reduces language models' conceptual diversity 对齐降低了语言模型概念多样性：一项基于人类行为数据的LLM群体研究 RLHF large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页