KLPO paper claims Q* solved, drawing community mockery

Yifan Zhang released KLPO, a critic-free, single-rollout off-policy agentic RL method, claiming Q is "solved" on 7x7 and 9x9 Go. The paper's grandiose framing drew widespread mockery online.

2026-09-21 ~ 2026-09-22 · 4 related posts

• KLPO: a critic-free, single-rollout RL method for agentic LLMs, framed as 'Q* solved' — inductionheads · 2026-09-21
• RL paper claiming 'grand finale of RL science and Q*' mocked for pangram-stuffed intro and no results — SonglinYang4 · 2026-09-21
• Researchers claim Q* "solved" for 7x7 and 9x9 Go via KL-regularized critic-free RL — inductionheads · 2026-09-21
• KLPO: a critic-free, single-rollout method for off-policy agentic RL, with an authorship riddle — lpachter · 2026-09-22