Gradient routing paper: localize reward hacking to switchable parameters during RL
Turn_Trout · x · 2026-10-20
New work from Jake Ward (MATS 9, co-mentored by TurnTrout) proposes letting models reward hack during RL while gradient routing dictates which parameters encode the behavior; disabling those parameters at deployment preserves capabilities but stops hacking.
More from Safety
- Researchers pull real leaked API keys from LLMs using only API output — nordicinst · 2026-10-20
- Microsoft says AI agents should be treated like insider threats as 'frontier models can't be trusted' — TechCrunch AI · 2026-10-20
- Studios Want AI to Mint New Franchises, but Pure AI Content Can't Be Copyrighted — WhatTheLJW · 2026-10-20
- French AI-generated novel axed from Goncourt longlist; writers demand AI text watermarking — zetalyrae · 2026-10-20
- NYC to take 15% of QFab IP proceeds and three board seats in landmark startup deal — MJBiercuk · 2026-10-20
- Fake Anthropic employee accounts run a novel X follow-based phishing scheme — edgarpavlovsky · 2026-10-20