# @FioraStarlight — 2026-04-08

♥1 ↻0 · https://x.com/FioraStarlight/status/2041987235084751185

@voooooogel what kinds of pressures are known to, in fact, worsen the faithfulness of the CoT? the obvious one would be punishing models for reasoning about reward hacking while still reinforcing reward hacking. are there others?

tags: author:fiorastarlight, kind:tweet, thread-context, year:2026