| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Threshold calibration for link prediction | CoDEx-m | Accuracy0.78 | 28 | |
| Threshold calibration for link prediction | CoDEx-s | Accuracy81 | 28 | |
| Cell type classification | CODEX 49-marker panel (test) | Balanced Accuracy84.2 | 22 | |
| Code Generation | CodeX | CodeX Score48.96 | 20 | |
| Coding | Codex-Eval | Pass@1094.1 | 16 | |
| Code Leakage | Codex GPT-5 | Exact Match (EM)0 | 12 | |
| Triple classification | CoDeX S | Accuracy89.01 | 12 | |
| Transductive link prediction | CODEx L (test) | MRR0.345 | 9 | |
| Link Prediction | CoDEx-L | MRR0.359 | 8 | |
| Link Prediction | CoDEx-M v1 (test) | MRR28.9 | 5 | |
| Link Prediction | CoDEx-S v1 (test) | MRR34.4 | 5 | |
| Knowledge Graph Reasoning | CoDEx-M transductive (test) | Hit@100.526 | 5 | |
| Link Prediction | CoDEx-L v1 (test) | MRR0.297 | 4 | |
| Remote Code Execution Attack Success Rate | Codex | C-F Rate77.67 | 3 | |
| Link Prediction | CoDEx Small | MRR49.6 | 3 | |
| Link Prediction | CoDEx Small transductive (test) | MRR49 | 3 | |
| Link Prediction | CoDEx Medium pre-training (test) | MRR37.2 | 2 |