English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Beyond F1: Evaluating Coverage and Failure Recovery in AI Model Security Scanners

Forum topic · 小凯 · 2026-08-30

Summary

A 2026 arXiv paper (2608.27424) by Lan et al. argues that conventional F1-based metrics only measure scanner accuracy when a usable security judgment is produced, ignoring coverage and failure modes. The authors benchmark three ML artifact scanners—ModelScan, ModelAudit, and Fickling—on a controlled corpus of 170 synthetic Pickle and PyTorch artifacts spanning 145 specimen families: 135 with binary security ground truth and 10 intentionally malformed without labels. They explicitly separate non-N/A coverage, analysis completion, definitive security decisions, non-security findings, and unsupported outcomes. On labeled families, ModelAudit produced definitive security decisions for all 135 families (100%), Fickling for 110 (81.5%), and ModelScan for 67 (49.6%). Conditioned on making a definitive judgment, ModelScan achieved 100% precision, recall, and F1—illustrating why judgment accuracy must be evaluated separately from judgment availability. The findings call for security evaluation frameworks that capture failure recovery and coverage, not just accuracy metrics.

Paper: Beyond F1: Evaluating Coverage and Failure Recovery in AI Model Security arXiv: 2608.27424 Authors: Qianlong Lan, Vinothini Pandurangan, Anuj Kaul, Indranil Sanyal Posted: 2026-08-27

Key Points

Static scanners are increasingly used to identify executable or otherwise unsafe content in machine-learning artifacts, yet conventional evaluation metrics characterize only cases where a scanner yields a usable security judgment.

The authors evaluate ModelScan, ModelAudit, and Fickling using a controlled, artifact-backed benchmark on a synthetic corpus of 170 Pickle and PyTorch focused artifacts across 145 specimen families — 135 with binary security ground truth and 10 intentionally malformed without labels.

They explicitly distinguish:

  • Non-N/A coverage
  • Analysis completion
  • Definitive security decisions
  • Non-security findings
  • Unsupported outcomes

Results on Labeled Families

| Scanner | Definitive security decisions | Coverage | |---|---|---| | ModelAudit | 135 / 135 | 100% | | Fickling | 110 / 135 | 81.5% | | ModelScan | 67 / 135 | 49.6% |

Conditioned on making a definitive judgment, ModelScan achieved 100% precision, recall, and F1 — demonstrating that judgment accuracy and judgment availability are orthogonal and must be evaluated separately.

The paper's central argument: security scanner evaluation should go beyond F1 to capture coverage and failure recovery behavior.

Tags

#machine-learning#security#arxiv#pickle#pytorch#static-analysis#benchmarking#evaluation-metrics

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178634236