SecGoal (Few-Shot): leaderboard
Metric: Extraction-Coverage F1 (0-1), the harmonic mean of extraction precision and property-level recall, for extracting formalizable security-goal statements from protocol documents with few-shot prompting, averaged over 15 protocols; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 0.5 |
| 2 | GLM-5 | 0.5 |
| 3 | Qwen 3.5 Plus | 0.49 |
| 4 | Gemini 3 Flash (Preview) | 0.43 |
| 5 | DeepSeek V4 Pro | 0.4 |
| 6 | DeepSeek V4 Flash | 0.4 |
| 7 | GPT-5.4 | 0.37 |
| 8 | Claude Sonnet 4.6 | 0.34 |
| 9 | Qwen 2.5 7B Instruct | 0.33 |
| 10 | Gemma 2 9B (IT) | 0.23 |
Interactive version: theaggregate.ai/benchmark?slug=secgoal-few-shot · How It Works · Data refreshed daily, snapshot 2026-10-07.