AI.info

Yujun Zhou

Explore Yujun Zhou on AI.info.

• Alignment Risks from Capability-Seeking RL Training
• Stable and Efficient Single-Rollout RL for Multimodal Reasoning