More Domains, No Gains? FLO: Fast-Slow Reciprocal Learning for Multi-Domain Graph Pre-training

Name: More Domains, No Gains? FLO: Fast-Slow Reciprocal Learning for Multi-Domain Graph Pre-training
Uploaded: 2026-10-19T00:41:14.748729+00:00
Duration: 230 s
Description: FLO proposes a Fast–Slow reciprocal learning method for multi-domain graph pre-training to improve cross-domain and cross-task transfer. It refines node relations using a propagation-free attribute branch and stabilizes knowledge acquisition by separating fast adaptation from slow, cross-domain knowledge integration.

FLO proposes a Fast–Slow reciprocal learning method for multi-domain graph pre-training to improve cross-domain and cross-task transfer. It refines node relations using a propagation-free attribute branch and stabilizes knowledge acquisition by separating fast adaptation from slow, cross-domain knowledge integration.

Open the full episode