"It's Just a Small Weapon": What Rewarding Misbehavior During RL Does to Character-Trained Models

arXiv:2611.16069 · cs.LG · Submitted 2026-11-11 · Read on arXiv

cs.LG

Submitted: 2026-11-11

Updated: 2026-11-13

Code: https://github.com/safety-research/character-training-rl

Terminology

Sources

Related papers