HNHacker News
TopNewBestAskShowJobs

categoricalrift

2 karma · joined October 30, 2025

submissionscomments
categoricalrift··on Autoresearch: Agents researching on single-GPU nanochat training automatically
How about the very last "Kept Improvement" in the plot? It's titled "random seed 42 -> 137". I do think this project is quite conceptually interesting, but the model literally choosing a different random seed to achieve lower loss feels pretty far removed from the flowery sci-fi writing at the top of the readme.