# Cheap Reward Hacking Detection

arXiv: https://arxiv.org/abs/2606.08893 · HTML: https://mlsanity.com/paper/2606.08893/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Reward hacking and misspecification, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 12 of its 20 references that are in the map.
- First posted on arXiv 8 Jun 2026.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Reward hacking and misspecification](https://mlsanity.com/alignment/reward-hacking-and-misspecification/)

## Builds on (most cited first, 12 of 12)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction](https://mlsanity.com/paper/1802.03426/) (2018)
- [Concrete Problems in AI Safety](https://mlsanity.com/paper/1606.06565/) (2016)
- [Towards Understanding Sycophancy in Language Models](https://mlsanity.com/paper/2310.13548/) (2023)
- [The Effects of Reward Misspecification: Mapping and Mitigating Misaligned Models](https://mlsanity.com/paper/2201.03544/) (2022)
- [FreeLB: Enhanced Adversarial Training for Natural Language Understanding](https://mlsanity.com/paper/1909.11764/) (2019)
- [SMART: Robust and Efficient Fine-Tuning for Pre-trained Natural Language Models through Principled Regularized Optimization](https://mlsanity.com/paper/1911.03437/) (2019)
- [Categorizing Variants of Goodhart's Law](https://mlsanity.com/paper/1803.04585/) (2018)
- [Reward Tampering Problems and Solutions in Reinforcement Learning: A Causal Influence Diagram Perspective](https://mlsanity.com/paper/1908.04734/) (2019)
- [MICo: Improved representations via sampling-based state similarity for Markov decision processes](https://mlsanity.com/paper/2106.08229/) (2021)
- [Terminal Wrench: A Dataset of 331 Reward-Hackable Environments and 3,632 Exploit Trajectories](https://mlsanity.com/paper/2604.17596/) (2026)

## Abstract (arXiv)

A small transformer encoder is trained to map Terminal-Wrench trajectories onto a unit sphere where embedding distance approximates the L1 distance between reward and metadata signals. A linear probe on top of that embedding detects reward hacking on the cleaned test split with AUC 0.9467 and TPR@5%FPR 0.8296, matching the TW sanitized LLM-as-judge AUC (0.9510 on the cleaned split) and exceeding its TPR@5%FPR (0.7130 vs 0.8296) on the same information condition, at roughly four orders of magnitude lower per-trajectory cost. The encoder is not a pure behavior reader: stripping natural-language reasoning from its input at probe time drops AUC to 0.6213.