Hacker News new | ask | show | jobs
Unlocking the Working Memory of Large Language Models for Latent Reasoning (arxiv.org)
2 points by korbip 23 days ago