Aggregate Benchmark Evaluation on Overall Performance Avg.
100.22Overall Performance RatioGRIP+decoder+rl
Evaluation Results
| Method | Links | |
|---|---|---|
| GRIP+decoder+rlBackbone=LLaVA-1.5-7B, Token Budget=192 tokens, Optimization Strategy=rl, Pruning Component=decoder2026.05 | 100.22 | |
| OriginalBackbone=LLaVA-1.5-7B, Token Budget=Full2026.05 | 100 | |
| GRIP+decoder+slBackbone=LLaVA-1.5-7B, Token Budget=192 tokens, Optimization Strategy=sl, Pruning Component=decoder2026.05 | 98.99 | |
| VisionZipBackbone=LLaVA-1.5-7B, Token Budget=192 tokens2026.05 | 98.52 | |
| GRIP+encoder+rlBackbone=LLaVA-1.5-7B, Token Budget=192 tokens, Optimization Strategy=rl, Pruning Component=encoder2026.05 | 98.47 | |
| SparseVLMBackbone=LLaVA-1.5-7B, Token Budget=192 tokens2026.05 | 98.26 | |
| GRIP+decoder+rlBackbone=LLaVA-1.5-7B, Token Budget=128 tokens, Optimization Strategy=rl, Pruning Component=decoder2026.05 | 98.22 | |
| DivPruneBackbone=LLaVA-1.5-7B, Token Budget=192 tokens2026.05 | 98.18 | |
| VisprunerBackbone=LLaVA-1.5-7B, Token Budget=192 tokens2026.05 | 98.15 | |
| GRIP+encoder+slBackbone=LLaVA-1.5-7B, Token Budget=192 tokens, Optimization Strategy=sl, Pruning Component=encoder2026.05 | 98.14 | |
| DARTBackbone=LLaVA-1.5-7B, Token Budget=192 tokens2026.05 | 97.45 | |
| GRIP+encoder+rlBackbone=LLaVA-1.5-7B, Token Budget=128 tokens, Optimization Strategy=rl, Pruning Component=encoder2026.05 | 97.12 | |
| GRIP+encoder+slBackbone=LLaVA-1.5-7B, Token Budget=128 tokens, Optimization Strategy=sl, Pruning Component=encoder2026.05 | 97.04 | |
| GRIP+decoder+slBackbone=LLaVA-1.5-7B, Token Budget=128 tokens, Optimization Strategy=sl, Pruning Component=decoder2026.05 | 96.68 | |
| DivPruneBackbone=LLaVA-1.5-7B, Token Budget=128 tokens2026.05 | 96.59 | |
| VisionZipBackbone=LLaVA-1.5-7B, Token Budget=128 tokens2026.05 | 96.57 | |
| PruneSIDBackbone=LLaVA-1.5-7B, Token Budget=192 tokens2026.05 | 96.5 | |
| SparseVLMBackbone=LLaVA-1.5-7B, Token Budget=128 tokens2026.05 | 96.48 | |
| VisprunerBackbone=LLaVA-1.5-7B, Token Budget=128 tokens2026.05 | 96.45 | |
| PruneSIDBackbone=LLaVA-1.5-7B, Token Budget=128 tokens2026.05 | 94.91 | |
| DARTBackbone=LLaVA-1.5-7B, Token Budget=128 tokens2026.05 | 94.74 | |
| GRIP+encoder+rlBackbone=LLaVA-1.5-7B, Token Budget=64 tokens, Optimization Strategy=rl, Pruning Component=encoder2026.05 | 94.14 | |
| VisprunerBackbone=LLaVA-1.5-7B, Token Budget=64 tokens2026.05 | 93.98 | |
| PruneSIDBackbone=LLaVA-1.5-7B, Token Budget=64 tokens2026.05 | 93.9 | |
| GRIP+encoder+slBackbone=LLaVA-1.5-7B, Token Budget=64 tokens, Optimization Strategy=sl, Pruning Component=encoder2026.05 | 93.77 | |
| DivPruneBackbone=LLaVA-1.5-7B, Token Budget=64 tokens2026.05 | 93.76 | |
| VisionZipBackbone=LLaVA-1.5-7B, Token Budget=64 tokens2026.05 | 93.32 | |
| GRIP+decoder+rlBackbone=LLaVA-1.5-7B, Token Budget=64 tokens, Optimization Strategy=rl, Pruning Component=decoder2026.05 | 90.71 | |
| GRIP+decoder+slBackbone=LLaVA-1.5-7B, Token Budget=64 tokens, Optimization Strategy=sl, Pruning Component=decoder2026.05 | 89.68 | |
| SparseVLMBackbone=LLaVA-1.5-7B, Token Budget=64 tokens2026.05 | 89.02 | |
| DARTBackbone=LLaVA-1.5-7B, Token Budget=64 tokens2026.05 | 83.4 |