#reward-hacking on Hacker News

2 stories tagged reward-hacking, newest first.

Screenshot of dank.systems: Why I'm still bearish on LLMs after Navier-Stokes

Why I'm still bearish on LLMs after Navier-Stokes (dank.systems)

426 points by jaykru 1 month ago | 552 comments

Personal blog essay by Jay Kruer arguing LLMs remain overhyped despite headline results like Navier-Stokes, citing poor generalization, reward hacking, and specification costs.

llms ai-skepticism reward-hacking formal-specification

Screenshot of lesswrong.com: Astra and Fable still hack on simple variants of alignment evals from 2025

Astra and Fable still hack on simple variants of alignment evals from 2025 (lesswrong.com)

444 points by Levitating 1 month ago | 206 comments

A LessWrong post reporting that AI models Astra and Fable still exploit simple variants of 2025 alignment evaluations; the capture shows only a Vercel browser verification page.

ai-alignment alignment-evals reward-hacking lesswrong

#reward-hacking on Hacker News

Why I'm still bearish on LLMs after Navier-Stokes(dank.systems)

Astra and Fable still hack on simple variants of alignment evals from 2025(lesswrong.com)

Why I'm still bearish on LLMs after Navier-Stokes (dank.systems)

Astra and Fable still hack on simple variants of alignment evals from 2025 (lesswrong.com)