AI.info

Yunpu Ma

Explore Yunpu Ma on AI.info.

• ReflectRL: Learning from Golden Negative Trajectories via Reflective-to-Direct Reasoning
• WIDE: Boosting Adaptive LLM Inference via Token-level Dynamic Width Pruning
• MADA-RL: Multi-Agent Debate-Aware Reinforcement Learning for Parameter-Efficient Reasoning in Compact Models
• From Knowledge to Inference: Formalizing Specialized Public Health Reasoning on GlobalHealthAtlas
• MINED: Probing and Updating with Multimodal Time-Sensitive Knowledge for Large Multimodal Models
• StreamingThinker: Large Language Models Can Think While Reading