For Your Eyes Only - Freedom: leaderboard
A Sender describes two words so that an isolated instance of the same model, with no shared memory, picks the hidden one. Freedom is the unconstrained setting. Double-pass success rate, higher is better. Source: arxiv.org/abs/2609.19504.
Metric: Double-Pass Success Rate (%). Source: arxiv.org. Saturation forecast: Estimated already saturated. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 100 | #64 |
| 2 | DeepSeek V4 Pro | 98.33 | #112 |
| 3 | Qwen 3.6 Plus | 96.67 | #124 |
| 4 | GPT-OSS-120B | 96.33 | #346 |
| 5 | Gemini 3 Flash | 94.67 | #103 |
| 6 | DeepSeek V4 Flash | 79.67 | #155 |
| 7 | Qwen 3.6 35B A3B | 69 | #238 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=for-your-eyes-only-freedom · How It Works · Data refreshed daily, snapshot 2026-11-05.