Syntactic Surprisal Update Measurement (KL-divergence) on SAP All (full)
38.5MVRRPT-1b
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| PT-1bFamily=Pythia2026.04 | 38.5 | 1.9 | 8.6 | 34.8 | 22.3 | |
| PT-160mFamily=Pythia2026.04 | 24.5 | 3.1 | 26.1 | 50.3 | 35.1 | |
| PT-2.8bFamily=Pythia2026.04 | 20.3 | 1.4 | 6.1 | 38 | 31.7 | |
| PT-70mFamily=Pythia2026.04 | 10.6 | 1.6 | 0.7 | 22.5 | 22.6 | |
| G2-smFamily=GPT-22026.04 | 10 | 0.4 | 1.7 | 11.2 | 5.3 | |
| PT-410mFamily=Pythia2026.04 | 9.9 | 1.2 | 0.6 | 9.2 | 26.7 | |
| G2-mdFamily=GPT-22026.04 | 9.7 | 0.8 | 2.6 | 18.9 | 20.9 | |
| OT-1.3bFamily=OPT2026.04 | 9.6 | 2.6 | 4.3 | 15.4 | 34.8 | |
| OT-2.7bFamily=OPT2026.04 | 9.5 | 2.1 | 4.4 | 11.5 | 34.3 | |
| OT-125mFamily=OPT2026.04 | 7.7 | 0.4 | 2.7 | 8.5 | 19.1 | |
| PT-1.4bFamily=Pythia2026.04 | 7.6 | 0.1 | 1.1 | 14.9 | 16 | |
| G2-lgFamily=GPT-22026.04 | 7.4 | 0.5 | 0.8 | 16 | 16.3 | |
| OT-13bFamily=OPT2026.04 | 6.5 | 1.7 | 3.8 | 17.5 | 26.4 | |
| OT-30bFamily=OPT2026.04 | 4.8 | 1.3 | 3.1 | 17 | 31.4 | |
| PT-12bFamily=Pythia2026.04 | 4.5 | 0.2 | 1.5 | 18.2 | 21.9 | |
| PT-6.9bFamily=Pythia2026.04 | 4.3 | 0.9 | 1.8 | 18.7 | 21.6 | |
| G2-xlFamily=GPT-22026.04 | 3.1 | 1 | 1.4 | 15.5 | 16.2 | |
| OT-6.7bFamily=OPT2026.04 | 2.7 | 0.8 | 1.1 | 10.2 | 24.6 | |
| OT-66bFamily=OPT2026.04 | 2.2 | 0.7 | 2.9 | 9.3 | 17.5 |