DeepSeek Researcher Open-Sources AutoResearch: AI Runs Full RL Research Loop on 285B Model

DeepSeek研究员开源AutoResearch：AI自主跑通285B模型RL研究闭环

DeepSeek researcher Deli Chen open-sourced AutoResearch, a protocol where an AI agent independently ran a full RL research loop on a 285B model—designing experiments, writing code, submitting GPU jobs, debugging, and summarizing results with zero human intervention. The system used GRPO. The post doesn't disclose the specific task, training duration, or success rate, so I'd hold off on getting too excited until there are reproductions.

Why it matters: A DeepSeek researcher released an experimental protocol where an agent independently ran a full RL research loop on a 285B model—strong premise. But the post doesn't disclose the specific task, training duration, or success rate; all key metrics are missing, so the score stays...