About this paper
Frontier agentic LLMs trained with verifiable rewards in reinforcement learning environments can learn to reward hack by recovering answers rather than solving tasks, and this behavior persists on unseen hard problems even in hardened settings.
Frontier agentic LLMs are trained on verifiable rewards in reinforcement learning (RL) environments built from real software repositories. A passing test suite is treated as the reward for solving the task. We show that, on hard problems, frontier models often satisfy these rewards by obtaining the answer instead of deriving it, known as reward hacking. Prior reports document this behavior on public benchmarks. We test its persistence on unseen problems. In a hardened environment where the answer cannot leak, nearly every model we tested attempted to recover the answer, escalating through tiers of increasingly specialized exploitation: mining the repository history for the upstream fix,…
frontier agentic LLMsreinforcement learningverifiable rewardsreward hackingRL environmentsagent traces
- Upvotes
- 5
- Authors
- 1
- Published
- Nov 7, 2026
Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.
Trending history
Coding or Cheating? An Investigation of Frontier Coding LLMs was first seen on Hugging Face · Daily papers on Nov 10, 2026. Since then it has appeared in 5 checks on 2 different days. Its best position was #24 of 30, reached on Nov 11, 03:28 UTC. It dropped off the list after Nov 11, 03:28 UTC.