List your company
ResearchFormerly WizWand

Multimodal Distribution Matching on 8-Gaussian synthetic landscape

0.001Last-epoch L1 DistanceBoosted GFlowNet (3 models, flow-additive)
0.0009470.0011690.001390.001611Nov 12, 2025
Updated 7mo ago

Evaluation results

MethodLinks
Boosted GFlowNet (3 models, flow-additive)Noise level (epsilon)=0.2, Sample size (n)=62025.110.001
Single GFlowNetNoise level (epsilon)=0.3, Sample size (n)=62025.110.001
Boosted GFlowNet (3 models, flow-additive)Noise level (epsilon)=0.3, Sample size (n)=62025.110.001
Boosted GFlowNet (2 models, flow-additive)Noise level (epsilon)=0.3, Sample size (n)=62025.110.0011
Single GFlowNetNoise level (epsilon)=0.2, Sample size (n)=62025.110.0012
Single GFlowNetNoise level (epsilon)=0.4, Sample size (n)=62025.110.0012
Boosted GFlowNet (2 models, flow-additive)Noise level (epsilon)=0.2, Sample size (n)=62025.110.0012
Boosted GFlowNet (3 models, flow-additive)Noise level (epsilon)=0.1, Sample size (n)=62025.110.0012
Boosted GFlowNet (3 models, flow-additive)Noise level (epsilon)=0.4, Sample size (n)=62025.110.0012
Boosted GFlowNet (2 models, flow-additive)Noise level (epsilon)=0.4, Sample size (n)=62025.110.0013
Boosted GFlowNet (3 models, flow-additive)Noise level (epsilon)=0, Sample size (n)=62025.110.0013
Boosted GFlowNet (2 models, flow-additive)Noise level (epsilon)=0.1, Sample size (n)=62025.110.0014
Boosted GFlowNet (3 models, flow-additive)Noise level (epsilon)=0.5, Sample size (n)=62025.110.0014
Boosted GFlowNet (2 models, flow-additive)Noise level (epsilon)=0.5, Sample size (n)=62025.110.0015
Boosted GFlowNet (2 models, flow-additive)Noise level (epsilon)=0, Sample size (n)=62025.110.0016
Single GFlowNetNoise level (epsilon)=0.5, Sample size (n)=62025.110.0017
Single GFlowNetNoise level (epsilon)=0, Sample size (n)=62025.110.0018
Single GFlowNetNoise level (epsilon)=0.1, Sample size (n)=62025.110.0018