List your company

Visual Question Answering on Text on TextVQA

58.21AccuracyOriginal
-1.60414413.92452829.453244.981872Mar 16, 2026Mar 25, 2026Apr 4, 2026Apr 14, 2026Apr 23, 2026May 3, 2026May 13, 2026

Evaluation results

MethodLinks
OriginalBackbone=LLaVA-1.5-7B, Token Budget=Full2026.0558.21
GRIP+decoder+slBackbone=LLaVA-1.5-7B, Token Budget=192 tokens, Optimization Strategy=sl, Pruning Component=decoder2026.0557.86
GRIP+decoder+rlBackbone=LLaVA-1.5-7B, Token Budget=192 tokens, Optimization Strategy=rl, Pruning Component=decoder2026.0557.79
SparseVLMBackbone=LLaVA-1.5-7B, Token Budget=192 tokens2026.0557.75
GRIP+encoder+slBackbone=LLaVA-1.5-7B, Token Budget=192 tokens, Optimization Strategy=sl, Pruning Component=encoder2026.0557.65
GRIP+encoder+rlBackbone=LLaVA-1.5-7B, Token Budget=192 tokens, Optimization Strategy=rl, Pruning Component=encoder2026.0557.6
VisprunerBackbone=LLaVA-1.5-7B, Token Budget=192 tokens2026.0557.45
VisionZipBackbone=LLaVA-1.5-7B, Token Budget=192 tokens2026.0557.33
GRIP+decoder+rlBackbone=LLaVA-1.5-7B, Token Budget=128 tokens, Optimization Strategy=rl, Pruning Component=decoder2026.0557.17
GRIP+encoder+slBackbone=LLaVA-1.5-7B, Token Budget=128 tokens, Optimization Strategy=sl, Pruning Component=encoder2026.0557
GRIP+encoder+rlBackbone=LLaVA-1.5-7B, Token Budget=128 tokens, Optimization Strategy=rl, Pruning Component=encoder2026.0556.93
DARTBackbone=LLaVA-1.5-7B, Token Budget=192 tokens2026.0556.9
GRIP+decoder+slBackbone=LLaVA-1.5-7B, Token Budget=128 tokens, Optimization Strategy=sl, Pruning Component=decoder2026.0556.87
VisionZipBackbone=LLaVA-1.5-7B, Token Budget=128 tokens2026.0556.83
VisprunerBackbone=LLaVA-1.5-7B, Token Budget=128 tokens2026.0556.8
SparseVLMBackbone=LLaVA-1.5-7B, Token Budget=128 tokens2026.0556.71
DivPruneBackbone=LLaVA-1.5-7B, Token Budget=192 tokens2026.0556.61
DivPruneBackbone=LLaVA-1.5-7B, Token Budget=128 tokens2026.0555.89
DARTBackbone=LLaVA-1.5-7B, Token Budget=128 tokens2026.0555.85
VisprunerBackbone=LLaVA-1.5-7B, Token Budget=64 tokens2026.0555.72
VisionZipBackbone=LLaVA-1.5-7B, Token Budget=64 tokens2026.0555.49
PruneSIDBackbone=LLaVA-1.5-7B, Token Budget=192 tokens2026.0555.07
GRIP+decoder+rlBackbone=LLaVA-1.5-7B, Token Budget=64 tokens, Optimization Strategy=rl, Pruning Component=decoder2026.0554.99
GRIP+encoder+rlBackbone=LLaVA-1.5-7B, Token Budget=64 tokens, Optimization Strategy=rl, Pruning Component=encoder2026.0554.79
PruneSIDBackbone=LLaVA-1.5-7B, Token Budget=64 tokens2026.0554.67
DivPruneBackbone=LLaVA-1.5-7B, Token Budget=64 tokens2026.0554.61
GRIP+encoder+slBackbone=LLaVA-1.5-7B, Token Budget=64 tokens, Optimization Strategy=sl, Pruning Component=encoder2026.0554.54
GRIP+decoder+slBackbone=LLaVA-1.5-7B, Token Budget=64 tokens, Optimization Strategy=sl, Pruning Component=decoder2026.0554.51
PruneSIDBackbone=LLaVA-1.5-7B, Token Budget=128 tokens2026.0554.45
SparseVLMBackbone=LLaVA-1.5-7B, Token Budget=64 tokens2026.0553.44
DARTBackbone=LLaVA-1.5-7B, Token Budget=64 tokens2026.0550.39
Qwen2.5-VL 7B, DenseBackbone=Qwen2.5-VL 7B, Sparsity=Dense2026.030.8293
ATV-PruningBackbone=Qwen2.5-VL 7B, Sparsity=60% Sparsity2026.030.7807
TAMPBackbone=Qwen2.5-VL 7B, Sparsity=60% Sparsity2026.030.7796
SparseGPTBackbone=Qwen2.5-VL 7B, Sparsity=60% Sparsity2026.030.7663
LLaVA-OneVision 7B, DenseBackbone=LLaVA-OneVision 7B, Sparsity=Dense2026.030.7605
WandaBackbone=Qwen2.5-VL 7B, Sparsity=60% Sparsity2026.030.756
SparseGPTBackbone=LLaVA-OneVision 7B, Sparsity=60% Sparsity2026.030.7262
ATV-PruningBackbone=LLaVA-OneVision 7B, Sparsity=60% Sparsity2026.030.7023
WandaBackbone=LLaVA-OneVision 7B, Sparsity=60% Sparsity2026.030.6964
TAMPBackbone=LLaVA-OneVision 7B, Sparsity=60% Sparsity2026.030.6964