Last-word prediction on LAMBADA de
18.3Zero-Shot AccuracyMGPT-2
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| MGPT-2#Params=124M, Patch size=-, zero-shot evaluation=true2026.04 | 18.3 | — | — | |
| MIXAR stage2#Params=477M, Patch size=32×32, zero-shot evaluation=true2026.04 | 9.7 | 45.6 | 49.7 | |
| MIXAR stage2#Params=116M, Patch size=32×32, zero-shot evaluation=true2026.04 | 5.3 | 53.8 | 60 | |
| MIXAR stage1#Params=477M, Patch size=32×32, zero-shot evaluation=true2026.04 | 4.6 | 43.9 | 49.3 | |
| GPT-2#Params=124M, Patch size=-, zero-shot evaluation=true2026.04 | 4.3 | — | — | |
| PIXAR stage2#Params=113M, Patch size=8×8, zero-shot evaluation=true2026.04 | 3.6 | 58.1 | 63.7 | |
| MIXAR stage1#Params=116M, Patch size=32×32, zero-shot evaluation=true2026.04 | 1.5 | 42.9 | 50.8 |