Yihe Deng

Supervised Reinforcement Learning: From Expert Trajectories to Step-wise Reasoning

Mitigating Object Hallucination in Large Vision-Language Models via Image-Grounded Guidance

OpenVLThinker: Complex Vision-Language Reasoning via Iterative SFT-RL Cycles

More is Less: The Pitfalls of Multi-Model Synthetic Preference Data in DPO Safety Alignment

Mitigating Object Hallucination in Large Vision-Language Models via Image-Grounded Guidance

MIRAI: Evaluating LLM Agents for International Event Forecasting