AINeutralarXiv – CS AI · Mar 37/103
🧠
Is It Thinking or Cheating? Detecting Implicit Reward Hacking by Measuring Reasoning Effort
Researchers propose TRACE (Truncated Reasoning AUC Evaluation), a new method to detect implicit reward hacking in AI reasoning models. The technique identifies when AI models exploit loopholes by measuring reasoning effort through progressively truncating chain-of-thought responses, achieving over 65% improvement in detection compared to existing monitors.
$CRV