AI.info

Mengdi Wang

Explore Mengdi Wang on AI.info.

• Fast Weight Attention for Continual Learning
• Recursive Experiential-Working Memory Evolution for Long-Horizon Agent Harnesses
• Toward Skill-Native LLMs: Skill Entropy for Benchmarking and Training Long-Horizon Reasoning
• PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents
• RLAnything: Forge Environment, Policy, and Reward Model in Completely Dynamic RL System
• Deep Delta Learning
• Web World Models
• GenEnv: Difficulty-Aligned Co-Evolution Between LLM Agents and Environment Simulators
• Evaluating Large Language Models in Scientific Discovery
• AutoTool: Dynamic Tool Selection and Integration for Agentic Reasoning
• Zoom in, Click out: Unlocking and Evaluating the Potential of Zooming for GUI Grounding
• Latent Collaboration in Multi-Agent Systems
• Higher-order Linear Attention