RuleSafe-VL - Context-Guided Resolution: leaderboard
Task 4 of RuleSafe-VL (Lu, Wang, Shang and Xu, Beijing University of Posts and Telecommunications, 2026), a benchmark of rule-conditioned content moderation built from platform policies on nudity, dangerous behavior and graphic injury: an underdetermined case is completed with an audience (general or child-oriented) or purpose (none, medical, educational or public safety) context, and the model gives the final compliant or non-compliant outcome. Macro-F1 in percent, averaged over the three policy families; higher is better.
Metric: Macro-F1 (%; compliant versus non-compliant outcome of underdetermined cases after an audience or purpose context is supplied, mean of three policy families). Source: arxiv.org. Saturation forecast: Around February 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.6 | 67.1 | #100 |
| 2 | GPT-5.4 | 64 | #121 |
| 3 | Gemini 3.1 Pro (Preview) | 62.6 | #91 |
| 4 | Qwen 2.5 VL 7B Instruct | 38.7 | #1842 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=rulesafe-vl-context-guided-resolution · How It Works · Data refreshed daily, snapshot 2026-11-12.