Training LLMs to Reason in a Continuous Latent Space by from Hacker News on 2024-12-10 16:26 (#6SV2X) Comments