cs.CL（2024-10-14）

📊 共 46 篇论文 | 🔗 10 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (41 🔗9) 支柱二：RL算法与架构 (RL & Architecture) (4 🔗1) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (41 篇)

#题目一句话要点标签🔗⭐
1 A Systematic Review on Prompt Engineering in Large Language Models for K-12 STEM Education 综述研究：大型语言模型结合提示工程在K-12 STEM教育中的应用与效果分析 large language model chain-of-thought
2 Balancing Continuous Pre-Training and Instruction Fine-Tuning: Optimizing Instruction-Following in LLMs 研究持续预训练与指令微调的平衡，优化LLM的指令遵循能力 large language model instruction following
3 SensorLLM: Aligning Large Language Models with Motion Sensors for Human Activity Recognition SensorLLM：通过传感器-语言对齐，赋能大语言模型进行人体活动识别 large language model foundation model ✅
4 Persistent Topological Features in Large Language Models 提出基于Zigzag持久同调的大语言模型层剪枝方法，保持系统整体视角。 large language model
5 Assessing Dialect Fairness and Robustness of Large Language Models in Reasoning Tasks 提出ReDial基准，评估大型语言模型在推理任务中对AAVE方言的公平性和鲁棒性。 large language model
6 Double Jeopardy and Climate Impact in the Use of Large Language Models: Socio-economic Disparities and Reduced Utility for Non-English Speakers 揭示大语言模型对非英语用户的双重不利影响 large language model
7 Rethinking Legal Judgement Prediction in a Realistic Scenario in the Era of Large Language Models 在真实场景下，利用大型语言模型重新审视法律判决预测 large language model
8 Skill Learning Using Process Mining for Large Language Model Plan Generation 融合过程挖掘技术，提升大语言模型生成复杂任务规划能力 large language model
9 Thinking LLMs: General Instruction Following with Thought Generation 提出一种无需额外人工数据的LLM训练方法，使其具备通用指令遵循的思考能力 instruction following
10 Gender Bias in Decision-Making with Large Language Models: A Study of Relationship Conflicts DeMET Prompts数据集揭示LLM在亲密关系决策中存在的性别偏见，安全措施可缓解 large language model
11 Denial-of-Service Poisoning Attacks against Large Language Models 提出基于投毒的拒绝服务攻击（P-DoS），突破LLM输出长度限制，提升攻击有效性。 large language model ✅
12 Large Language Models Are Active Critics in NLG Evaluation 提出Active-Critic，使LLM在NLG评估中从被动遵循转为主动适应。 large language model
13 Large Language Model Evaluation via Matrix Nuclear-Norm 提出矩阵核范数以高效评估大型语言模型的压缩能力 large language model ✅
14 MentalGLM Series: Explainable Large Language Models for Mental Health Analysis on Chinese Social Media 提出MentalGLM系列模型，用于中文社交媒体心理健康分析的可解释大语言模型。 large language model ✅
15 A Comparative Study of Translation Bias and Accuracy in Multilingual Large Language Models for Cross-Language Claim Verification 研究多语言LLM在跨语言声明验证中的翻译偏差与准确性，揭示低资源语言的性能瓶颈。 large language model
16 EffiCoder: Enhancing Code Generation in Large Language Models through Efficiency-Aware Fine-tuning EffiCoder：通过效率感知微调增强大型语言模型的代码生成能力 large language model ✅
17 RoCoFT: Efficient Finetuning of Large Language Models with Row-Column Updates RoCoFT：通过行列更新高效微调大型语言模型 large language model
18 Generative AI and Its Impact on Personalized Intelligent Tutoring Systems 生成式AI赋能个性化智能辅导系统，提升教育效果与公平性 large language model multimodal
19 Augmenting In-Context-Learning in LLMs via Automatic Data Labeling and Refinement 提出ADLR自动标注与优化方法，提升LLM在复杂推理任务中的上下文学习能力 large language model chain-of-thought
20 Minimum Tuning to Unlock Long Output from LLMs with High Quality Data as the Key 通过高质量数据微调，以低成本解锁LLM的长文本生成能力 large language model foundation model
21 LLM Unlearning via Loss Adjustment with Only Forget Data 提出FLAT方法，仅用遗忘数据调整损失，实现大语言模型高效解学习。 large language model
22 Your Mixture-of-Experts LLM Is Secretly an Embedding Model For Free 无需微调！MoE LLM的专家路由权重可作为即用型嵌入模型 large language model
23 Mix Data or Merge Models? Optimizing for Diverse Multi-Task Learning 提出多语言多任务学习中基于目标和语言的模型融合方法，提升安全性和通用性能。 large language model
24 Not All Options Are Created Equal: Textual Option Weighting for Token-Efficient LLM-Based Knowledge Tracing 提出LOKT框架，通过文本选项加权提升LLM在知识追踪中的效率和可解释性。 large language model
25 On Calibration of LLM-based Guard Models for Reliable Content Moderation 评估并校准LLM守卫模型，提升内容审核的可靠性 large language model
26 Towards Acyclic Preference Evaluation of Language Models via Multiple Evaluators 提出PGED框架，通过多评估器集成解决语言模型偏好评估中的循环矛盾问题 large language model
27 PRACTIQ: A Practical Conversational Text-to-SQL dataset with Ambiguous and Unanswerable Queries PRACTIQ：构建包含歧义和无法回答问题的实用对话式文本到SQL数据集 large language model
28 Assessing Bias in Metric Models for LLM Open-Ended Generation Bias Benchmarks 评估LLM开放生成偏差基准中度量模型的偏见 large language model
29 LLMs know their vulnerabilities: Uncover Safety Gaps through Natural Distribution Shifts 提出ActorBreaker方法，揭示LLM在自然分布偏移下的安全漏洞 large language model ✅
30 Jailbreak Instruction-Tuned LLMs via end-of-sentence MLP Re-weighting 提出基于末句MLP重加权的白盒攻击方法，破解指令微调LLM的安全机制 large language model
31 Beyond-RAG: Question Identification and Answer Generation in Real-Time Conversations 提出超越RAG的实时对话问答系统，提升客服效率并降低运营成本 large language model
32 Active Learning for Robust and Representative LLM Generation in Safety-Critical Scenarios 提出基于主动学习与聚类的框架，提升LLM在安全场景下的生成质量与代表性 large language model
33 DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Heads DuoAttention：利用检索头和流式头实现高效长文本LLM推理 large language model ✅
34 Use Random Selection for Now: Investigation of Few-Shot Selection Strategies in LLM-based Text Augmentation for Classification 研究表明，在基于LLM的文本增强分类任务中，随机样本选择策略通常优于更复杂的选择策略。 large language model
35 Recipe for Zero-shot POS Tagging: Is It Useful in Realistic Scenarios? 针对低资源场景，研究零样本POS标注的有效数据集选择策略 large language model
36 Medico: Towards Hallucination Detection and Correction with Multi-source Evidence Fusion Medico：融合多源证据的大语言模型幻觉检测与纠正框架 large language model
37 Parenting: Optimizing Knowledge Selection of Retrieval-Augmented Language Models with Parameter Decoupling and Tailored Tuning Parenting框架通过解耦参数空间优化RAG中知识选择，提升模型可靠性。 large language model
38 A Unified Approach to Routing and Cascading for LLMs 提出统一的级联路由框架，优化LLM的成本-性能权衡 large language model
39 Locking Down the Finetuned LLMs Safety 提出SafetyLock，通过激活向量干预提升微调LLM的安全性 large language model ✅
40 FunnelRAG: A Coarse-to-Fine Progressive Retrieval Paradigm for RAG 提出FunnelRAG，一种由粗到精的渐进式检索范式，提升RAG效率。 large language model
41 AlphaLoRA: Assigning LoRA Experts Based on Layer Training Quality AlphaLoRA：基于层训练质量分配LoRA专家，提升大模型微调效率。 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🔗⭐
42 How to Leverage Demonstration Data in Alignment for Large Language Model? A Self-Imitation Learning Perspective 提出GSIL框架，利用自模仿学习高效对齐大语言模型与离线演示数据 imitation learning large language model instruction following
43 Improving the Language Understanding Capabilities of Large Language Models Using Reinforcement Learning 利用强化学习提升大型语言模型在自然语言理解任务上的能力 reinforcement learning PPO large language model
44 Ada-K Routing: Boosting the Efficiency of MoE-based LLMs 提出Ada-K路由，通过动态调整专家激活数量提升MoE-LLM效率。 PPO large language model
45 Temperature-Centric Investigation of Speculative Decoding with Knowledge Distillation 研究温度对推测解码的影响，提出知识蒸馏在一致温度下的应用以加速高温度生成。 distillation ✅

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
46 MMCFND: Multimodal Multilingual Caption-aware Fake News Detection for Low-resource Indic Languages 提出MMCFND框架，用于低资源印度语多模态假新闻检测，并构建了多语言数据集MMIFND。 manipulation multimodal

⬅️ 返回 cs.CL 首页 · 🏠 返回主页