cs.AI（2024-10-03）

📊 共 13 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) (9) 支柱二：RL算法与架构 (RL & Architecture) (3 🔗1) 支柱一：机器人控制 (Robot Control) (1 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (9 篇)

#题目一句话要点标签🔗⭐
1 Plots Unlock Time-Series Understanding in Multimodal Models 利用时序图解锁多模态模型对时间序列数据的理解 foundation model multimodal
2 IoT-LLM: a framework for enhancing Large Language Model reasoning from real-world sensor data IoT-LLM框架通过融合物联网传感器数据增强大语言模型的物理世界推理能力 large language model chain-of-thought
3 From Pixels to Tokens: Byte-Pair Encoding on Quantized Visual Modalities 提出基于BPE的图像Token化方法，提升多模态大语言模型对视觉信息的理解能力。 large language model foundation model multimodal
4 AI-rays: Exploring Bias in the Gaze of AI Through a Multimodal Interactive Installation AI-rays：通过多模态交互装置探索AI凝视中的偏见 multimodal
5 Cognitive Biases in Large Language Models for News Recommendation 研究大型语言模型在新闻推荐中的认知偏差，并提出缓解策略。 large language model
6 The Role of Deductive and Inductive Reasoning in Large Language Models 提出DID方法，通过动态融合演绎推理和归纳推理，提升大语言模型在复杂推理任务中的性能。 large language model
7 Large Language Models as Markov Chains 将大语言模型等价于马尔可夫链，从而分析其泛化能力。 large language model
8 Can Large Language Models Grasp Legal Theories? Enhance Legal Reasoning with Insights from Multi-Agent Collaboration 提出MALR框架，利用多Agent协作提升LLM在复杂法律推理中的表现 large language model
9 LLM Safeguard is a Double-Edged Sword: Exploiting False Positives for Denial-of-Service Attacks 提出利用假阳性进行拒绝服务攻击的新方法 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (3 篇)

#题目一句话要点标签🔗⭐
10 From Imitation to Exploration: End-to-end Autonomous Driving based on World Model RAMBLE：基于世界模型的端到端自动驾驶，融合模仿学习与强化学习 reinforcement learning imitation learning world model ✅
11 SEAL: SEmantic-Augmented Imitation Learning via Language Model SEAL：通过语言模型增强语义的模仿学习，解决长时决策任务。 imitation learning large language model
12 LLaMA-Berry: Pairwise Optimization for O1-like Olympiad-Level Mathematical Reasoning 提出LLaMA-Berry以解决大型语言模型的数学推理能力不足问题 reinforcement learning RLHF large language model

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
13 CAnDOIT: Causal Discovery with Observational and Interventional Data from Time-Series CAnDOIT：利用观测和干预时序数据进行因果发现，适用于复杂机器人环境 manipulation ✅

⬅️ 返回 cs.AI 首页 · 🏠 返回主页