SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2307.01646 · 2023

SwinGNN: Rethinking Permutation Invariance in Diffusion Models for Graph Generation

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 12 functions out of this paper's own repositories and ran 4 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
qiyan98/swingnn canonical 3 of 9
dsl-lab/swingnn canonical 1 of 3
FunctionStatusWhere it lives
window_partition Ran qiyan98/swingnn/model/swin_gnn/swin_gnn.py
code served (permissive licence) · get_code("144d10b49baeb8a6")
citeseer_ego Ran dsl-lab/swingnn/setup/gen_graph_data.py
code served (permissive licence) · get_code("33e65ffda4746402")
weight_init Ran qiyan98/swingnn/model/unet/unet_edm.py
code served (permissive licence) · get_code("d41a4250066bce93")
window_reverse Ran qiyan98/swingnn/model/swin_gnn/swin_gnn.py
code served (permissive licence) · get_code("61bf152e6a42a184")
clustering_worker Not yet run qiyan98/swingnn/evaluation/stats.py
code served (permissive licence) · get_code("85b7ab6da324d0ed")
compute_score Not yet run qiyan98/swingnn/runner/sanity_check_helper.py
code served (permissive licence) · get_code("632d673f64dd5dcc")
construct_atomic_number_array Not yet run qiyan98/swingnn/setup/mol_preprocess.py
code served (permissive licence) · get_code("0c1e757e3c9f7161")
degree_worker Not yet run qiyan98/swingnn/evaluation/stats.py
code served (permissive licence) · get_code("36afd1c572ae4673")
gen_graph_comm_grid_lobster Not yet run dsl-lab/swingnn/setup/gen_graph_data.py
code served (permissive licence) · get_code("eefc727475cbfb4d")
get_all_permutation Not yet run qiyan98/swingnn/runner/sanity_check_helper.py
code served (permissive licence) · get_code("c20bf9ed5347eaa1")
n_community Not yet run dsl-lab/swingnn/setup/gen_graph_data.py
code served (permissive licence) · get_code("d520133b61f52af3")
pad_array Not yet run qiyan98/swingnn/evaluation/mmd.py
code served (permissive licence) · get_code("d1af971a6d642b82")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Diffusion models based on permutation-equivariant networks can learn permutation-invariant distributions for graph data. However, in comparison to their non-invariant counterparts, we have found that these invariant models encounter greater learning challenges since 1) their effective target distributions exhibit more modes; 2) their optimal one-step denoising scores are the score functions of Gaussian mixtures with more components. Motivated by this analysis, we propose a non-invariant diffusion model, called $\textit{SwinGNN}$, which employs an efficient edge-to-edge 2-WL message passing network and utilizes shifted window based self-attention inspired by SwinTransformers. Further, through systematic ablations, we identify several critical training and sampling techniques that significantly improve the sample quality of graph generation. At last, we introduce a simple post-processing trick, $\textit{i.e.}$, randomly permuting the generated graphs, which provably converts any graph generative model to a permutation-invariant one. Extensive experiments on synthetic and real-world protein and molecule datasets show that our SwinGNN achieves state-of-the-art performances. Our code is released at https://github.com/qiyan98/SwinGNN.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2307.01646")
get_code_for_paper("2307.01646")
have("2307.01646")

Connect an agent — have() is free.