mlsanityData to 14 Oct 2026

Against Proxy Optimization

arXiv 2606.23597 ↗First posted 22 Jun 2026cs.AIcited by 1

Not on the map. Why?

Abstract (arXiv)

I discuss conditions under which maximizing a proxy utility function is harmful and suggest this poses problems for applying decision theory.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 15 of its 49 references that are in the map. First posted on arXiv 22 Jun 2026.

Builds on

15 of 49 references.

References · 15

  1. 2022Scaling Laws for Reward Model Overoptimizationcited by 1,095
  2. 2025Reasoning Models Don't Always Say What They Thinkcited by 415
  3. 2019Risks from Learned Optimization in Advanced Machine Learning Systemscited by 258
  4. 2023Reward Model Ensembles Help Mitigate Overoptimizationcited by 231
  5. 2022Defining and Characterizing Reward Hackingcited by 203
  6. 2018Categorizing Variants of Goodhart's Lawcited by 149
  7. 2019Reward Tampering Problems and Solutions in Reinforcement Learning: A Causal Influence Diagram Perspectivecited by 108
  8. 2016The Off-Switch Gamecited by 97
  9. 2021Consequences of Misaligned AIcited by 80
  10. 2024Feedback Loops With Language Models Drive In-Context Reward Hackingcited by 77
  11. 2024Correlated Proxies: A New Definition and Improved Mitigation for Reward Hackingcited by 50
  12. 2024On Targeted Manipulation and Deception when Optimizing LLMs for User Feedbackcited by 47
3 more
  1. 2023Goodhart's Law in Reinforcement Learningcited by 41
  2. 2020The Problem with Metrics is a Fundamental Problem for AIcited by 37
  3. 2025Off-Switching Not Guaranteedcited by 4

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Fragility of Value under Imperfect Alignmentcited by 0