AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Sequence-to-Set Generative Models

Longtao Tang, Ying Zhou, Yu Yang

NeurIPS 2022 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Score-based generative modeling analysis · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add size-bias trick (ablation)Benchmark · TMALLaverage l1-distance: 0.95 -> 0.91Improved
add size-bias trick (ablation)DCM · TMALLaverage l1-distance: 1.00 -> 0.94Improved
add size-bias trick (ablation)MRW · TMALLaverage l1-distance: 1.02 -> 0.92Improved
add size-bias trick (ablation)SetNN · TMALLaverage l1-distance: 0.96 -> 0.88Improved
add size-bias trick (ablation)GRU2Set · TMALLaverage l1-distance: 0.97 -> 0.87Improved
add size-bias trick (ablation)Benchmark · HKTVMALLaverage l1-distance: 0.83 -> 0.80Improved
add size-bias trick (ablation)DCM · HKTVMALLaverage l1-distance: 0.86 -> 0.81Improved
add size-bias trick (ablation)MRW · HKTVMALLaverage l1-distance: 0.88 -> 0.82Improved
add size-bias trick (ablation)SetNN · HKTVMALLaverage l1-distance: 0.88 -> 0.79Improved
add size-bias trick (ablation)GRU2Set · HKTVMALLaverage l1-distance: 0.88 -> 0.78Improved
use DCM or MRW instead of histogram benchmarkDCM, MRW · TMALL, HKTVMALLl1-distance: 0.95 -> 1.00Hurt
GRU2Set with size bias vs benchmarkGRU2Set + Size Bias · TMALLl1-distance: 4%Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy