LangFIR: Efficient Language Steering Using Monolingual Data
Discover LangFIR, a method that uses sparse features from monolingual data to improve language steering in large language models with high accuracy.
How Incentives Boost Creativity with Generative AI
Discover how incentive structures impact human-AI collaboration, enhancing creativity and diversity in generative AI-driven tasks.
User-Centric Salient Object Detection Beyond Visual Cues
Discover User Salient Object Detection (UserSOD), a novel approach that aligns object detection with user intent for improved image analysis results.
Optimizing Neurorobot Learning with Limited Demo Data
Enhance neurorobot policies using preference regret to learn effectively from limited demonstration data and improve robotic control performance.
Source-Code Taxonomy of Coding Agent Architectures
Explore a detailed source-code taxonomy of coding agent architectures, analyzing control strategies, tools, and context management in LLM-based agents.