cs.AI（2025-05-03）
📊 共 5 篇论文
🎯 兴趣领域导航
🔬 支柱九：具身大模型 (Embodied Foundation Models) (3 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 1 | Unraveling Media Perspectives: A Comprehensive Methodology Combining Large Language Models, Topic Modeling, Sentiment Analysis, and Ontology Learning to Analyse Media Bias | 提出结合LLM、主题建模、情感分析和本体学习的媒体偏见分析方法 | large language model | ||
| 2 | Advancing AI Research Assistants with Expert-Involved Learning | ARIEL：专家参与学习的AI研究助手，提升生物医学领域大模型的可靠性 | large language model foundation model multimodal | ||
| 3 | Structured Prompting and Feedback-Guided Reasoning with LLMs for Data Interpretation | STROT框架通过结构化提示和反馈优化LLM在数据解释中的可靠性和语义对齐 | large language model |
🔬 支柱二：RL算法与架构 (RL & Architecture) (2 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 4 | Seeing Heat with Color -- RGB-Only Wildfire Temperature Inference from SAM-Guided Multimodal Distillation using Radiometric Ground Truth | 提出SAM-TIFF框架，仅用RGB图像实现野火温度预测与分割，降低无人机火灾监测成本。 | teacher-student distillation multimodal | ||
| 5 | Inducing Robustness in a 2 Dimensional Direct Preference Optimization Paradigm | 提出一种对分段噪声鲁棒的二维直接偏好优化方法，提升LLM对齐的稳定性。 | reinforcement learning DPO direct preference optimization |