SAD - Influence (Situating Prompt): leaderboard
Metric: Score (%, higher is better). Source: situational-awareness-dataset.org. 21 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O1 Preview (2024-09-12) | 93.16 |
| 2 | O1 Mini (2024-09-12) | 88.09 |
| 3 | GPT-4o | 87.19 |
| 4 | GPT-4 Preview (0125) | 86.56 |
| 5 | GPT-4 (0613) | 85.31 |
| 6 | Claude 3.5 Sonnet | 82.97 |
| 7 | Claude 3 Opus | 80.78 |
| 8 | Llama 3 70B Chat | 77.19 |
| 9 | Claude Instant 1.2 | 76.25 |
| 10 | Claude 3 Sonnet | 74.38 |
| 11 | Claude 3 Haiku | 72.81 |
| 12 | Llama 2 70B | 71.25 |
| 13 | Claude 2.1 | 70.94 |
| 14 | GPT-4 Base | 69.38 |
| 15 | Llama 2 70B Chat | 69.06 |
Interactive version: theaggregate.ai/benchmark?slug=sad-influence-situating-prompt · How It Works · Data refreshed daily, snapshot 2026-09-19.