A hypothetical benchmark threshold; it does not imply clinical fitness or medical approval.
Current Yes probability
51%
+4.0 pts24h
| Date | Yes probability |
|---|---|
| Jul 4 | 32% |
| Jul 5 | 31% |
| Jul 6 | 29% |
| Jul 7 | 31% |
| Jul 8 | 35% |
| Jul 9 | 37% |
| Jul 10 | 36% |
| Jul 11 | 35% |
| Jul 12 | 38% |
| Jul 13 | 41% |
| Jul 14 | 42% |
| Jul 15 | 41% |
| Jul 16 | 41% |
| Jul 17 | 44% |
| Jul 18 | 47% |
| Jul 19 | 47% |
| Jul 20 | 46% |
| Jul 21 | 48% |
| Jul 22 | 50% |
| Jul 23 | 51% |
Resolution criteria
Resolves Yes if reproducible third-party evaluation of a publicly downloadable model exceeds 90% on the frozen benchmark.
Evidence source
Simulated independent model leaderboard
Adjacent signals