Anyone else hacking on long-horizon reasoning frameworks for LLMs?github.com/Mehditerranean282 points89Mehdi9 months ago