Hacker News new | ask | show | jobs
by dataviz1000 1 hour ago
Reinforcement learning can solve a Rubik’s Cube. A LLM that hasn’t been trained to solve a Rubik’s Cube can not.