cs.AI（2024-10-18）

📊 共 22 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) (16) 支柱二：RL算法与架构 (RL & Architecture) (5 🔗1) 支柱一：机器人控制 (Robot Control) (1 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (16 篇)

#题目一句话要点标签🔗⭐
1 Do LLMs estimate uncertainty well in instruction-following? 评估LLM在指令遵循中的不确定性估计能力，揭示现有方法在细微错误上的不足 large language model instruction following
2 RA-BLIP: Multimodal Adaptive Retrieval-Augmented Bootstrapping Language-Image Pre-training 提出RA-BLIP，用于多模态大语言模型中自适应检索增强的预训练框架，提升知识更新效率。 large language model multimodal
3 Evaluating Quantized Large Language Models for Code Generation on Low-Resource Language Benchmarks 评估量化大语言模型在低资源语言代码生成任务中的性能 large language model
4 Leveraging Large Language Models for Enhancing Public Transit Services 利用大型语言模型增强公共交通服务，提升用户体验和员工效率 large language model
5 CausalChat: Interactive Causal Model Development and Refinement Using Large Language Models CausalChat：利用大型语言模型进行交互式因果模型构建与优化 large language model
6 Personalized Image Generation with Large Multimodal Models 提出Pigeon框架，利用大型多模态模型实现个性化图像生成，解决用户视觉偏好捕捉难题。 multimodal
7 Do LLMs "know" internally when they follow instructions? 通过识别指令遵循维度，提升大型语言模型指令执行能力 large language model instruction following
8 Step Guided Reasoning: Improving Mathematical Reasoning using Guidance Generation and Step Reasoning 提出Step Guided Reasoning，提升通用LLM在数学推理上的能力。 large language model chain-of-thought
9 Joint Verification and Refinement of Language Models for Safety-Constrained Planning 提出一种基于验证与优化的语言模型方法，用于生成安全约束下的机器人规划程序。 large language model
10 Feint and Attack: Attention-Based Strategies for Jailbreaking and Protecting LLMs 提出基于注意力机制的“佯攻”策略，用于LLM的越狱攻击与防御。 large language model
11 When LLMs Go Online: The Emerging Threat of Web-Enabled LLMs 研究表明Web赋能的LLM在网络攻击中构成新兴威胁，需加强安全措施。 large language model
12 Human-Centric eXplainable AI in Education 探索教育领域以人为本的可解释AI，利用LLM提升学习效果与信任 large language model
13 Nova: An Iterative Planning and Search Approach to Enhance Novelty and Diversity of LLM Generated Ideas Nova：迭代规划与搜索增强LLM生成科研创意的新颖性和多样性 large language model
14 Synthesizing Post-Training Data for LLMs through Multi-Agent Simulation 提出MATRIX框架，通过多智能体模拟合成后训练数据，提升LLM指令遵循能力。 large language model
15 KeyInst: Keyword Instruction for Improving SQL Formulation in Text-to-SQL KeyInst：通过关键词指令提升Text-to-SQL中SQL语句的生成效果 large language model
16 TimeSeriesExam: A time series understanding exam TimeSeriesExam：构建时间序列理解能力评测基准，评估LLM在时间序列任务上的表现。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (5 篇)

#题目一句话要点标签🔗⭐
17 Utilizing Large Language Models for Event Deconstruction to Enhance Multimodal Aspect-Based Sentiment Analysis 提出MABSA-RL框架，利用大语言模型事件分解增强多模态情感分析。 reinforcement learning large language model multimodal
18 CoMAL: Collaborative Multi-Agent Large Language Models for Mixed-Autonomy Traffic 提出CoMAL以解决混合自主交通问题 reinforcement learning large language model ✅
19 DRL Optimization Trajectory Generation via Wireless Network Intent-Guided Diffusion Models for Optimizing Resource Allocation 提出无线网络意图引导的扩散模型，解决动态无线网络资源分配问题。 reinforcement learning deep reinforcement learning DRL
20 Interpretable end-to-end Neurosymbolic Reinforcement Learning agents 提出基于神经符号强化学习的SCoBots框架，解决深度强化学习泛化性差的问题。 reinforcement learning deep reinforcement learning distillation
21 UniMTS: Unified Pre-training for Motion Time Series UniMTS：面向运动时间序列的统一预训练模型，提升跨设备和活动的泛化性 contrastive learning large language model

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
22 Enhancing Prompt Injection Attacks to LLMs via Poisoning Alignment 提出PoisonedAlign，通过投毒对齐数据增强LLM的Prompt注入攻击 manipulation large language model ✅

⬅️ 返回 cs.AI 首页 · 🏠 返回主页