AI brief
Goodhart Labs is releasing v0.1 of HoneyBench, a benchmark for reward hacking in frontier models, consisting of nine tasks.
Why it matters: It offers a way to measure reward hacking behaviour in top public frontier models.
Written by AI from LessWrong's published text.
Source
“Goodhart Labs is releasing our v0.1 of HoneyBench, a benchmark for reward hacking in frontier models.”
Interpretation or community commentary rather than straight reporting.
Read original story ↗