RTL Delay Prediction on DeepRTL2 1.0 (test)
0.773R2 ScoreDeepRTL2
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| DeepRTL2Backbone=Llama, Hard Negatives=Included2025.05 | 0.773 | 46.9 | 0.476 | |
| DeepRTL2Backbone=DeepSeek, Hard Negatives=Included2025.05 | 0.772 | 44.8 | 0.478 | |
| DeepRTL2Backbone=DeepSeek, Hard Negatives=Excluded2025.05 | 0.743 | 44.9 | 0.507 | |
| DeepRTL2Backbone=Llama, Hard Negatives=Excluded2025.05 | 0.735 | 47.1 | 0.515 | |
| text-embedding-3-large2025.05 | 0.699 | 70.5 | 0.548 | |
| GritLM-7B2025.05 | 0.651 | 72.6 | 0.591 | |
| text-embedding-3-small2025.05 | 0.608 | 88.3 | 0.626 |