LINKED LIST
[txt mode]
▸
Measuring Reward-Seeking by Instillin...
home
explore
|
log in
Measuring Reward-Seeking by Instilling Contrastive Beliefs
alignment.openai.com · 2026-09-25 · 0 upvotes
log in
to save, upvote or flag this.
─── In 0 lists ─────────────────────────────────────────
(not in any lists yet)
related read on:
*
Astra and Fable still hack on simple variants of alignment evals from 2025 — LessWrong
(from the discussion)
*
METR and Redwood Offer Holy #%^@ Postmortem Of The HuggingFace Hack
(from the discussion)
*
The Rise and Fall of Agent Civilizations
(from the discussion)