Research map / Probabilistic and scientific modeling / Variational and generative sampling
Exchangeable Generative Models with Flow Scans
Christopher Bender, Kevin O'Connor, Yang Li, Juan Garcia, Junier Oliva, Manzil Zaheer
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use FlowScan instead of the strongest reported baseline, NS | Exchangeable density estimation · Synthetic | PPLL: -1.07 -> 0.14 | Improved |
| Use FlowScan instead of the strongest reported baseline, NS | Exchangeable density estimation · Airplanes | PPLL: 4.09 -> 4.81 | Improved |
| Use FlowScan instead of the strongest reported baseline, NS | Exchangeable density estimation · Chairs | PPLL: 2.02 -> 2.58 | Improved |
| Use FlowScan instead of the strongest reported baseline, NS | Exchangeable density estimation · ModelNet10 | PPLL: 2.12 -> 3.01 | Improved |
| Use FlowScan instead of the strongest reported baseline, NS | Exchangeable density estimation · ModelNet10a | PPLL: 2.82 -> 3.58 | Improved |
| Use FlowScan instead of the strongest reported baseline, NS | Exchangeable density estimation · Caudate | PPLL: 4.49 -> 4.87 | Improved |
| Use FlowScan instead of the strongest reported baseline, NS | Exchangeable density estimation · Thalamus | PPLL: 2.69 -> 3.12 | Improved |
| Use FlowScan instead of the strongest reported baseline, NS | Exchangeable density estimation · SpatialMNIST | PPLL: -5.37 -> -5.26 | Improved |
| Add convolution-based Set-Coupling to BRUNO while retaining the i.i.d. base likelihood | MNIST image sets · MNIST | PPLL: -643.6 -> -634.8 | Improved |
| Add sorting, scan likelihood, and 3d convolution-based correspondence coupling | Set-Coupling model · MNIST | PPLL: -634.8 -> -621.7 | Improved |
| Add correspondence coupling (ablation) | FlowScan on ModelNet10 · ModelNet10 | PPLL: 2.67 -> 3.01 | Improved |
| Add equivariant flow transformation (ablation) | FlowScan on ModelNet10 · ModelNet10 | PPLL: 2.34 -> 2.67 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- BRUNO — BRUNO: A Deep Recurrent Model for Exchangeable Data (NeurIPS 2018) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Tractable Density Estimation on Learned Manifolds with Conformal Embedding Flows (NeurIPS 2021 · similarity 0.87)
- C-Flow: Conditional Generative Flow Models for Images and 3D Point Clouds (CVPR 2020 · similarity 0.87)
- Residual Flows for Invertible Generative Modeling (NeurIPS 2019 · similarity 0.86)
- Expanding Flow Maps (NeurIPS 2026 · similarity 0.85)
- Beckmann Transport Models: From Autonomous Flows to One-Step Maps (NeurIPS 2026 · similarity 0.85)
- Not-So-Optimal Transport Flows for 3D Point Cloud Generation (ICLR 2025 · similarity 0.85)
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.
