LightDiC: A Simple yet Effective Approach for Large-scale Digraph Representation Learning
Summary: LightDiC: scalable digraph convolution via the magnetic Laplacian, moves topology work to offline preprocessing so downstream training is non‑recursive and efficient at large scale. Proves complex-field message passing ≈ proximal gradient descent on Dirichlet energy (digraph denoising), yielding strong expressiveness; matches or beats SOTA with fewer parameters. (summarized by gpt-5-mini on Feb 09 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Xunkai Li
- 2. Meihao Liao
- 3. Zhengyu Wu
- 4. Daohan Su
- 5. Wentao Zhang
- 6. Rong-Hua Li
- 7. Guoren Wang
Incoming Citations (Sorted by Pagerank)
Showing 1 of 1 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 10,545 | OpenFGL: A Comprehensive Benchmark for Federated Graph Learning | 2025 | VLDB | 4.1945683e-05 |
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 0 of 0 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next