# benedekrozemberczki/CapsGNN

A PyTorch implementation of "Capsule Graph Neural Network" (ICLR 2019).

Repository: https://github.com/benedekrozemberczki/CapsGNN
Canonical: https://ross.abutalabs.com/products/capsgnn
Language: Python
License: GPL-3.0
License Family: copyleft
Topics: capsgnn, capsule-network, capsule-neural-networks, convolution, pytorch, research, tensorflow, machine-learning, deep-learning, gnn, deepwalk, node2vec, graph-attention-networks, graph-attention-model, graph-classification, sklearn, struc2vec, graph-convolution, graph-neural-network
Last push: 2023-03-18T12:08:55+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2773, "days_push": 1264, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1278, forks 197 (observed 2026-08-28T04:04:13.345444+00:00)

## What it is
A PyTorch implementation of the Capsule Graph Neural Network (CapsGNN) from ICLR 2019, which uses capsule networks and routing mechanisms to generate graph embeddings for graph classification. It is research code accompanying the published paper.

## Use cases
- classify graphs with a capsule graph neural network
- reproduce CapsGNN results from the ICLR 2019 paper
- generate graph-level embeddings from node features
- benchmark graph classification models on standard datasets
- study capsule networks applied to graph-structured data

## When to choose
- you need a reference implementation of CapsGNN for research or reproduction
- you want to experiment with capsule-based graph classification in PyTorch

## When to avoid
- you need a production-ready or actively maintained GNN framework
- you want broad model support beyond the CapsGNN architecture
- you need compatibility with recent PyTorch versions without adaptation

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning
- domain: machine-learning, deep-learning, graph-processing
- platform: python
- tags: graph-neural-network, capsule-network, graph-classification, pytorch, iclr-2019, research-code

## Member repositories
- benedekrozemberczki/CapsGNN (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:13.345444+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T05:02:44.496472+00:00, confidence not recorded.
  - readme: https://github.com/benedekrozemberczki/CapsGNN (fetched 2026-08-28T04:04:13.345444+00:00, sha daa82bfb6cd3)
- Data as of 2026-08-30T08:39:29.467469+00:00.