Vision-Language Modeling on VisP
42AccuracyZPPO
Evaluation results
| Method | Links | |
|---|---|---|
| ZPPOModel Scale=2B, Prompt Replay Buffer Augmentation=true2026.06 | 42 | |
| GRPOModel Scale=2B, Prompt Replay Buffer Augmentation=true2026.06 | 40.3 | |
| GRPOModel Scale=2B, Prompt Replay Buffer Augmentation=false2026.06 | 39.2 | |
| On-DistillModel Scale=2B, Prompt Replay Buffer Augmentation=false2026.06 | 38.8 | |
| On-DistillModel Scale=2B, Prompt Replay Buffer Augmentation=true2026.06 | 38.8 | |
| Off-DistillModel Scale=2B, Prompt Replay Buffer Augmentation=false2026.06 | 38.4 | |
| Qwen3.5Model Scale=2B, Prompt Replay Buffer Augmentation=false2026.06 | 38.3 | |
| Off-DistillModel Scale=2B, Prompt Replay Buffer Augmentation=true2026.06 | 37.8 | |
| ZPPOModel Scale=0.8B, Prompt Replay Buffer Augmentation=true2026.06 | 35 | |
| GRPOModel Scale=0.8B, Prompt Replay Buffer Augmentation=true2026.06 | 27.8 | |
| GRPOModel Scale=0.8B, Prompt Replay Buffer Augmentation=false2026.06 | 24.2 | |
| On-DistillModel Scale=0.8B, Prompt Replay Buffer Augmentation=false2026.06 | 21.2 | |
| On-DistillModel Scale=0.8B, Prompt Replay Buffer Augmentation=true2026.06 | 20.9 | |
| Off-DistillModel Scale=0.8B, Prompt Replay Buffer Augmentation=true2026.06 | 20.8 | |
| Qwen3.5Model Scale=0.8B, Prompt Replay Buffer Augmentation=false2026.06 | 20.5 | |
| Off-DistillModel Scale=0.8B, Prompt Replay Buffer Augmentation=false2026.06 | 20 |