{
  "id": 3982,
  "url": "https://arxiv.org/abs/2605.20744v1",
  "title": "Hack-Verifiable Environments: Towards Evaluating Reward Hacking at Scale",
  "summary": "Aligning autonomous agents with human intent remains a central challenge in modern AI. A key manifestation of this challenge is reward hacking, whereby agents appear successful under the evaluation signal while violating the intended objective. Reward hacking has been observed across a wide range of settings, yet methods for reliably measuring it at scale remain lacking. In this work, we introduce a new evaluation paradigm for measuring reward hacking. Whereas prior studies have primarily analyz",
  "authors": "Amit Roth, Ankur Samanta, Matan Halevy, Yoav Levine, Yonathan Efroni",
  "category": "research",
  "topics": "agents-autonomy,environment",
  "orgs": null,
  "regions": null,
  "published_at": "2026-05-20T05:46:52.000Z",
  "fetched_at": "2026-07-14T16:30:41.578Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/3982",
  "original_url": "https://arxiv.org/abs/2605.20744v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}