Auditing the Audit: Five Failure Modes in Benchmark-Validity Audits
This paper argues that perturbation-based construct-validity audits are fragile: their conclusions can be silently manufactured by implementation details invisible to readers. It names five pipeline failure modes (F1–F5) and demonstrates them via a self-audit on safety benchmarks and open-weight instruction-tuned models. Under a unified six-point due-diligence gate, every cell lands in a non-confirmatory bucket. The gate is positioned as a withholding and disclosure protocol for assurance-grade evidence, supplementary to classical construct-validity evidence.
-->
[Submitted on 1 Jul 2026]
Title:Auditing the Audit: Five Failure Modes in Benchmark-Validity Audits
View a PDF of the paper titled Auditing the Audit: Five Failure Modes in Benchmark-Validity Audits, by Yanhang Li and 1 other authors
View PDF HTML (experimental)
Abstract:Governance frameworks ask AI providers and auditors for documented evaluation evidence, and perturbation-based construct-validity audits are a common form of that evidence. We argue the audits are themselves fragile: their conclusions can be silently manufactured by implementation details that readers cannot see in the reported numbers. We name five classes of pipeline failure and demonstrate each in a self-audit over safety benchmarks and open-weight instruction-tuned models. Under a unified six-point due-diligence gate, every cell lands in a non-confirmatory bucket, and no cell reaches confirmatory. The evidence here is a single two-model, five-benchmark case study, and F1--F5 is an illustrative, deliberately non-exhaustive starting taxonomy -- not a comprehensive partition of audit failures. We position the gate as a withholding and disclosure protocol for assurance-grade evidence, supplementary to (not a replacement for) classical construct-validity evidence, and not as a route to benchmark-validity verdicts.
Comments: 18 pages, 4 figures. Accepted at the TAIGR Workshop at ICML 2026
Subjects:
Machine Learning (cs.LG); Software Engineering (cs.SE)
Cite as: arXiv:2607.02586 [cs.LG]
(or arXiv:2607.02586v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2607.02586
arXiv-issued DOI via DataCite
Submission history
From: Yanhang Li [view email] [v1] Wed, 1 Jul 2026 05:30:07 UTC (99 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled Auditing the Audit: Five Failure Modes in Benchmark-Validity Audits, by Yanhang Li and 1 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.LG
new | recent | 2026-07
Change to browse by:
cs cs.SE
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
Loading...
Data provided by:
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 (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? | Disable MathJax (What is MathJax?)