Acceptance Cards:A Four-Diagnostic Standard for Safe Fine-Tuning Defense Claims
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913112663261184 |
|---|---|
| author | Konrad, Phongsakon Mark Tanyel, Toygar Ayvaz, Serkan |
| author_facet | Konrad, Phongsakon Mark Tanyel, Toygar Ayvaz, Serkan |
| contents | Safe fine-tuning defenses are often endorsed on the basis of a held-out gap reduction, but the same reduction can come from sampling noise, subject artifacts, capability loss, or a mechanism that does not transfer. We introduce Acceptance Cards: an evaluation protocol, a documentation object, an executable audit package, and a claim-specific evidential standard for safe fine-tuning defense claims. The protocol checks statistical reliability, fresh semantic generalization, mechanism alignment, and cross-task transfer before treating a gap reduction as a full-card pass. Re-scored under this installed-gap protocol, SafeLoRA fails the full-card pass on Gemma-2-2B-it: under strict mechanism-class coding it fails all four diagnostics, and under a permissive shrinkage relabel it still fails three of four. This is a narrow installed-gap audit on one model family, not a global judgment of SafeLoRA's effectiveness. In a 46-cell audit, no cell satisfies the strict conjunction. The closest family is a near miss that passes reliability and mechanism checks where the required data are available, but fails the fresh-subject threshold, lacks a strict transfer pass, and carries a measurable deployment-accuracy cost. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_10575 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Acceptance Cards:A Four-Diagnostic Standard for Safe Fine-Tuning Defense Claims Konrad, Phongsakon Mark Tanyel, Toygar Ayvaz, Serkan Cryptography and Security Artificial Intelligence Machine Learning Safe fine-tuning defenses are often endorsed on the basis of a held-out gap reduction, but the same reduction can come from sampling noise, subject artifacts, capability loss, or a mechanism that does not transfer. We introduce Acceptance Cards: an evaluation protocol, a documentation object, an executable audit package, and a claim-specific evidential standard for safe fine-tuning defense claims. The protocol checks statistical reliability, fresh semantic generalization, mechanism alignment, and cross-task transfer before treating a gap reduction as a full-card pass. Re-scored under this installed-gap protocol, SafeLoRA fails the full-card pass on Gemma-2-2B-it: under strict mechanism-class coding it fails all four diagnostics, and under a permissive shrinkage relabel it still fails three of four. This is a narrow installed-gap audit on one model family, not a global judgment of SafeLoRA's effectiveness. In a 46-cell audit, no cell satisfies the strict conjunction. The closest family is a near miss that passes reliability and mechanism checks where the required data are available, but fails the fresh-subject threshold, lacks a strict transfer pass, and carries a measurable deployment-accuracy cost. |
| title | Acceptance Cards:A Four-Diagnostic Standard for Safe Fine-Tuning Defense Claims |
| topic | Cryptography and Security Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2605.10575 |