logoalt Hacker News

dataviz1000today at 8:11 PM0 repliesview on HN

Reinforcement learning can solve a Rubik’s Cube. A LLM that hasn’t been trained to solve a Rubik’s Cube can not.