Skip to content

Comment on Project Vend: Phase Twoparent

Comments

It's likely that the weaknesses have a shared foundation: LLM pre-training fails to teach those LLMs to be good at agentic behavior, creating a lasting deficiency.

No known way to fully solve that as of yet, but, as always, we can mitigate with better training. Modern RLVR-trained LLMs are already much better at tasks like this than they were a year ago.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.