The Perils of Optimizing Learned Reward Functions: Low Training Error Does Not Guarantee Low Regret

Lukas Fluri, Leon Lang, Alessandro Abate, Patrick Forré, David Krueger +1 more
2/14/2026

No confident code implementation found for this paper.

Semantic Scholar

Code Implementations

No confident code match yet

We couldn't find an author-owned or strongly-evidenced community implementation for this paper. Any repos shown below are weak matches — verify before relying on them.

No code implementations found yet.

Know of an implementation? Let us know in the comments below!

Cite this paper

@article{fluri2026perils,
  title  = {The Perils of Optimizing Learned Reward Functions: Low Training Error Does Not Guarantee Low Regret},
  author = {Lukas Fluri and Leon Lang and Alessandro Abate and Patrick Forré and David Krueger and J. Skalse},
  year   = {2026},
  url    = {https://api.semanticscholar.org/CorpusID:87d9a1bb8f9f74aecfd21c22000459842293e1c9},
  journal = {ICML 2025 2025}
}

Discussion