[2606.08893] Cheap Reward Hacking Detection

-->

Computer Science > Machine Learning

arXiv:2606.08893 (cs)

[Submitted on 8 Jun 2026]

Title:Cheap Reward Hacking Detection

Authors:Iván Belenky, Joaquín Itria, Steven Johns<br>View a PDF of the paper titled Cheap Reward Hacking Detection, by Iv\'an Belenky and Joaqu\'in Itria and Steven Johns

View PDF<br>HTML (experimental)

Abstract:A small transformer encoder is trained to map Terminal-Wrench trajectories onto a unit sphere where embedding distance approximates the $L_1$ 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 $0.6213$.

Comments:<br>20 pages, 6 figures, 12 tables

Subjects:

Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)

ACM classes:<br>I.2.6; I.2.7

Cite as:<br>arXiv:2606.08893 [cs.LG]

(or<br>arXiv:2606.08893v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2606.08893

Focus to learn more

arXiv-issued DOI via DataCite (pending registration)

Submission history<br>From: Ivan Belenky [view email]<br>[v1]<br>Mon, 8 Jun 2026 00:35:54 UTC (822 KB)

Full-text links:<br>Access Paper:

View a PDF of the paper titled Cheap Reward Hacking Detection, by Iv\'an Belenky and Joaqu\'in Itria and Steven Johns<br>View PDF<br>HTML (experimental)<br>TeX Source

view license

Current browse context:

cs.LG

next >

new<br>recent<br>| 2026-06

Change to browse by:

cs<br>cs.AI<br>cs.CR

References & Citations

NASA ADS<br>Google Scholar

Semantic Scholar

export BibTeX citation<br>Loading...

BibTeX formatted citation

&times;

loading...

Data provided by:

Bookmark

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

IArxiv recommender toggle

IArxiv Recommender<br>(What is IArxiv?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs .

Which authors of this paper are endorsers? |<br>Disable MathJax (What is MathJax?)