ThreatCore (SRL Prompting): leaderboard
ThreatCore threat classification (explicit, implicit or none; Bruni, Bardazzi and Tesconi, 2026) with semantic-role prompting: the model first extracts the actor, action and victim of any harm and then classifies the text with those roles in the prompt. Macro-F1 over the three classes in percent, mean of five runs; higher is better.
Metric: Macro-F1 (%; explicit threat, implicit threat and non-threat, mean of five runs; the model first extracts the actor, action and victim roles and classifies with them in the prompt). Source: arxiv.org. Saturation forecast: Around February 2027. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-OSS-20B | 68.7 | #1150 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=threatcore-srl-prompting · How It Works · Data refreshed daily, snapshot 2026-11-12.