PARD: Permutation-invariantAutoregressiveDiffusion forGraphGeneration
–Neural Information Processing Systems
Specifically, we show that contrary to sets, elements in a graph are not entirely unordered and there is a unique partial order for nodes and edges. With this partial order,PARD generates a graph in a block-by-block, autoregressivefashion, where each block'sprobability isconditionally modeled by a shared diffusion model with an equivariant network.
Neural Information Processing Systems
Feb-7-2026, 20:53:21 GMT
- Country:
- Asia > Japan
- Honshū > Tōhoku > Miyagi Prefecture > Sendai (0.04)
- North America > United States
- Pennsylvania > Allegheny County > Pittsburgh (0.04)
- Asia > Japan
- Genre:
- Research Report (0.46)
- Technology: