Formal Verification on Memory Scheduler
29Assertion CountSaarthi
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SaarthiBackbone=GPT-5, Pass level=Pass@32026.03 | 29 | — | 1 | 32.01 | 48.86 | |
| SaarthiBackbone=GPT-5, Pass level=Pass@22026.03 | 28 | — | 0 | 32.14 | 58.21 | |
| SaarthiBackbone=GPT-4.1, Pass level=Pass@12026.03 | 24 | — | 2 | 35.71 | 42.93 | |
| SaarthiBackbone=GPT-4.1, Pass level=Pass@32026.03 | 23 | — | 1 | 21.74 | 46.39 | |
| SaarthiBackbone=GPT-5, Pass level=Pass@12026.03 | 22 | — | 2 | 40.91 | 54.73 | |
| SaarthiBackbone=Llama3.3, Pass level=Pass@32026.03 | 17 | — | 0 | 17.65 | 37.3 | |
| SaarthiBackbone=GPT-4.1, Pass level=Pass@22026.03 | 16 | — | 0 | 50 | 43.32 | |
| SaarthiBackbone=Llama3.3, Pass level=Pass@12026.03 | 13 | — | 3 | 30.77 | 39.47 | |
| SaarthiBackbone=Llama3.3, Pass level=Pass@22026.03 | 7 | — | 0 | 71.43 | 44.2 |