List your company
ResearchFormerly WizWand

Image Classification on Stanford Cars (Top-1/Top-5 Accuracy)

92.9Top-1 AccuracyBiMamba
0.308824.346948.38572.4231Mar 27, 2026Apr 4, 2026Apr 12, 2026Apr 21, 2026Apr 29, 2026May 7, 2026May 16, 2026
Updated 5mo ago

Evaluation results

MethodLinks
BiMambaScale=Tiny, Params (M)=7.72026.0592.9—
BiMambaScale=Base, Params (M)=89.92026.0592.5—
BiLSTMScale=Tiny, Params (M)=7.72026.0592.3—
BiLSTMScale=Base, Params (M)=83.22026.0592.3—
BiMambaScale=Small, Params (M)=26.92026.0592.1—
BiLSTMScale=Small, Params (M)=23.92026.0592.1—
AttnScale=Base, Params (M)=86.52026.0592.1—
AttnScale=Small, Params (M)=22.92026.0591.7—
AttnScale=Tiny, Params (M)=5.72026.0590.5—
JACTUSBackbone=ViT-Large, Budget=80%2026.0587.8—
JACTUSBackbone=ViT-Large, Budget=60%2026.0586.9—
PiSSABackbone=ViT-Large, Budget=100%2026.0586.7—
SVFTBackbone=ViT-Large, Budget=100%2026.0584.9—
JACTUSBackbone=ViT-Base, Budget=80%2026.0584.7—
DoRABackbone=ViT-Large, Budget=100%2026.0584.2—
LoRABackbone=ViT-Large, Budget=100%2026.0583.3—
SVD-LLM+LoRABackbone=ViT-Large, Budget=80%2026.0582.6—
DoRABackbone=ViT-Base, Budget=100%2026.0581.8—
SVD-LLM+LoRABackbone=ViT-Large, Budget=60%2026.0581.3—
PiSSABackbone=ViT-Base, Budget=100%2026.0579.1—
JACTUSBackbone=ViT-Base, Budget=60%2026.0578.9—
Fine-tuned (STL)Backbone=CLIP-ViT-B/32, Bit-width=Full-precision, Training=Single-task learning2026.0578.3—
JACTUSBackbone=ViT-Large, Budget=40%2026.0577.7—
SVFTBackbone=ViT-Base, Budget=100%2026.0577.5—
LoRABackbone=ViT-Base, Budget=100%2026.0577.1—
SVD-LLM+LoRABackbone=ViT-Base, Budget=80%2026.0576.8—
SVD-LLM+LoRABackbone=ViT-Large, Budget=40%2026.0572.5—
WUDI-MergingBackbone=CLIP-ViT-B/32, Merging Strategy=WUDI-Merging, Bit-width=Full-precision2026.0572.5—
SVD-LLM+LoRABackbone=ViT-Base, Budget=60%2026.0571.5—
WUDI-Merging w/ GPTQBackbone=CLIP-ViT-B/32, Merging Strategy=WUDI-Merging, Quantization Algorithm=GPTQ, Bit-width=4-bit2026.0568.9—
WUDI-Merging w/ E-PMQBackbone=CLIP-ViT-B/32, Merging Strategy=WUDI-Merging, Quantization Algorithm=E-PMQ, Bit-width=4-bit2026.0568.9—
Simple Averaging w/ E-PMQBackbone=CLIP-ViT-B/32, Merging Strategy=Simple Averaging, Quantization Algorithm=E-PMQ, Bit-width=4-bit2026.0567.8—
JACTUSBackbone=ViT-Base, Budget=40%2026.0567—
TIES-Merging w/ E-PMQBackbone=CLIP-ViT-B/32, Merging Strategy=TIES-Merging, Quantization Algorithm=E-PMQ, Bit-width=4-bit, Lambda=0.32026.0566.7—
WUDI-Merging w/ E-PMQMerging Strategy=WUDI-Merging, Quantization Method=E-PMQ, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0565.63—
WUDI-MergingMerging Strategy=WUDI-Merging, Quantization Method=None, Model Backbone=CLIP-ViT-B/32, Bit-width=Full2026.0564.8—
Task Arithmetic w/ E-PMQBackbone=CLIP-ViT-B/32, Merging Strategy=Task Arithmetic, Quantization Algorithm=E-PMQ, Bit-width=4-bit, Lambda=0.32026.0564.4—
TIES-MergingBackbone=CLIP-ViT-B/32, Merging Strategy=TIES-Merging, Bit-width=Full-precision, Lambda=0.32026.0564.2—
Simple Averaging w/ E-PMQMerging Strategy=Simple Averaging, Quantization Method=E-PMQ, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0563.95—
WUDI-Merging w/ RTNBackbone=CLIP-ViT-B/32, Merging Strategy=WUDI-Merging, Quantization Algorithm=RTN, Bit-width=4-bit2026.0563.8—
WUDI-Merging w/ AWQBackbone=CLIP-ViT-B/32, Merging Strategy=WUDI-Merging, Quantization Algorithm=AWQ, Bit-width=4-bit2026.0563.8—
TIES-Merging w/ E-PMQMerging Strategy=TIES-Merging, Quantization Method=E-PMQ, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0563.26—
WUDI-Merging w/ GPTQMerging Strategy=WUDI-Merging, Quantization Method=GPTQ, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0562.77—
Simple AveragingBackbone=CLIP-ViT-B/32, Merging Strategy=Simple Averaging, Bit-width=Full-precision2026.0562.6—
TIES-Merging w/ GPTQBackbone=CLIP-ViT-B/32, Merging Strategy=TIES-Merging, Quantization Algorithm=GPTQ, Bit-width=4-bit, Lambda=0.32026.0561.1—
Simple Averaging + E-PMQMerging Strategy=Simple Averaging, Quantization Method=E-PMQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0560.63—
Simple AveragingMerging Strategy=Simple Averaging, Quantization Method=None, Model Backbone=CLIP-ViT-B/32, Bit-width=Full2026.0560.4—
WUDI-Merging w/ AWQMerging Strategy=WUDI-Merging, Quantization Method=AWQ, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0560.09—
Pre-trainedBackbone=CLIP-ViT-B/32, Bit-width=Full-precision2026.0559.8—
Simple AveragingMerging Strategy=Simple Averaging, Quantization Method=Full-precision, Bit-width=Full, Backbone=CLIP-ViT-B/322026.0559.6—
Simple Averaging w/ GPTQBackbone=CLIP-ViT-B/32, Merging Strategy=Simple Averaging, Quantization Algorithm=GPTQ, Bit-width=4-bit2026.0559.5—
TIES-Merging + E-PMQMerging Strategy=TIES-Merging, Quantization Method=E-PMQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0558.85—
Task Arithmetic w/ E-PMQMerging Strategy=Task Arithmetic, Quantization Method=E-PMQ, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0558.48—
Simple Averaging w/ GPTQMerging Strategy=Simple Averaging, Quantization Method=GPTQ, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0558.34—
WUDI-Merging w/ RTNMerging Strategy=WUDI-Merging, Quantization Method=RTN, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0557.8—
WUDI-Merging + E-PMQMerging Strategy=WUDI-Merging, Quantization Method=E-PMQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0557.41—
Simple Averaging + GPTQMerging Strategy=Simple Averaging, Quantization Method=GPTQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0557.07—
Task ArithmeticBackbone=CLIP-ViT-B/32, Merging Strategy=Task Arithmetic, Bit-width=Full-precision, Lambda=0.32026.0555.7—
TIES-Merging w/ AWQBackbone=CLIP-ViT-B/32, Merging Strategy=TIES-Merging, Quantization Algorithm=AWQ, Bit-width=4-bit, Lambda=0.32026.0555.4—
TIES-Merging w/ RTNBackbone=CLIP-ViT-B/32, Merging Strategy=TIES-Merging, Quantization Algorithm=RTN, Bit-width=4-bit, Lambda=0.32026.0555—
Simple Averaging w/ RTNBackbone=CLIP-ViT-B/32, Merging Strategy=Simple Averaging, Quantization Algorithm=RTN, Bit-width=4-bit2026.0554.7—
TIES-MergingMerging Strategy=TIES-Merging, Quantization Method=None, Model Backbone=CLIP-ViT-B/32, Bit-width=Full2026.0554.6—
SVD-LLM+LoRABackbone=ViT-Base, Budget=40%2026.0554.5—
Simple Averaging w/ RTNMerging Strategy=Simple Averaging, Quantization Method=RTN, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0553.75—
Task Arithmetic w/ GPTQBackbone=CLIP-ViT-B/32, Merging Strategy=Task Arithmetic, Quantization Algorithm=GPTQ, Bit-width=4-bit, Lambda=0.32026.0553.5—
TIES-Merging w/ GPTQMerging Strategy=TIES-Merging, Quantization Method=GPTQ, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0553.17—
Simple Averaging + RTNMerging Strategy=Simple Averaging, Quantization Method=RTN, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0553.12—
Task Arithmetic + E-PMQMerging Strategy=Task Arithmetic, Quantization Method=E-PMQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0552.39—
Simple Averaging w/ AWQBackbone=CLIP-ViT-B/32, Merging Strategy=Simple Averaging, Quantization Algorithm=AWQ, Bit-width=4-bit2026.0551.7—
Simple Averaging w/ AWQMerging Strategy=Simple Averaging, Quantization Method=AWQ, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0551—
Simple Averaging + AWQMerging Strategy=Simple Averaging, Quantization Method=AWQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0550.58—
TIES-Merging w/ AWQMerging Strategy=TIES-Merging, Quantization Method=AWQ, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0548.09—
Task Arithmetic w/ AWQBackbone=CLIP-ViT-B/32, Merging Strategy=Task Arithmetic, Quantization Algorithm=AWQ, Bit-width=4-bit, Lambda=0.32026.0546.3—
TIES-Merging w/ RTNMerging Strategy=TIES-Merging, Quantization Method=RTN, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0545.94—
Task Arithmetic w/ RTNBackbone=CLIP-ViT-B/32, Merging Strategy=Task Arithmetic, Quantization Algorithm=RTN, Bit-width=4-bit, Lambda=0.32026.0545.7—
WUDI-MergingMerging Strategy=WUDI-Merging, Quantization Method=Full-precision, Bit-width=Full, Backbone=CLIP-ViT-B/322026.0544.8—
WUDI-Merging + GPTQMerging Strategy=WUDI-Merging, Quantization Method=GPTQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0542.31—
WUDI-Merging + AWQMerging Strategy=WUDI-Merging, Quantization Method=AWQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0537.56—
WUDI-Merging + RTNMerging Strategy=WUDI-Merging, Quantization Method=RTN, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0537.5—
TIES-MergingMerging Strategy=TIES-Merging, Quantization Method=Full-precision, Bit-width=Full, Backbone=CLIP-ViT-B/322026.0536.2—
TIES-Merging + GPTQMerging Strategy=TIES-Merging, Quantization Method=GPTQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0535.22—
Task ArithmeticMerging Strategy=Task Arithmetic, Quantization Method=None, Model Backbone=CLIP-ViT-B/32, Bit-width=Full2026.0533.2—
TIES-Merging + AWQMerging Strategy=TIES-Merging, Quantization Method=AWQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0531.75—
Task Arithmetic w/ GPTQMerging Strategy=Task Arithmetic, Quantization Method=GPTQ, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0531.05—
TIES-Merging + RTNMerging Strategy=TIES-Merging, Quantization Method=RTN, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0530.15—
Task Arithmetic w/ AWQMerging Strategy=Task Arithmetic, Quantization Method=AWQ, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0527.16—
Task Arithmetic w/ RTNMerging Strategy=Task Arithmetic, Quantization Method=RTN, Model Backbone=CLIP-ViT-B/32, Bit-width=4-bit2026.0524.9—
Task ArithmeticMerging Strategy=Task Arithmetic, Quantization Method=Full-precision, Bit-width=Full, Backbone=CLIP-ViT-B/322026.0512.2—
Task Arithmetic + GPTQMerging Strategy=Task Arithmetic, Quantization Method=GPTQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.0511.65—
Task Arithmetic + AWQMerging Strategy=Task Arithmetic, Quantization Method=AWQ, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.059.66—
Task Arithmetic + RTNMerging Strategy=Task Arithmetic, Quantization Method=RTN, Bit-width=4-bit, Backbone=CLIP-ViT-B/322026.058.57—
ProxyNCALoss function=Proxy Neighborhood Component Analysis2026.036.48—
CE+SILLoss function=Cross-Entropy + Silhouette2026.035.9616.63
CE+SIL+SupCon2Loss function=Cross-Entropy + Silhouette + SupCon22026.035.8515.91
SupCon2Loss function=Supervised Contrastive (two-view)2026.035.5615.66
CELoss function=Cross-Entropy2026.034.9115.05
SupConLoss function=Supervised Contrastive2026.034.8415.78
CenterLoss function=Center Loss2026.033.8713.84