FiLM conditioning targeted at early message-passing layers lets pretrained GNS models generalize to new material properties using only 12 trajectories, a 5-fold data reduction versus multi-task baselines.
Title resolution pending
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.LG 3verdicts
UNVERDICTED 3representative citing papers
GGDA framework generates knowledge-preserving intermediate graphs via FGW metric and a vertex-based progression to enable gradual domain adaptation across large graph distribution shifts.
DIB-OD isolates a stable invariant core in heterogeneous graph representations via orthogonal subspace decomposition, IB teacher-student distillation, HSIC independence, and confidence-gated regularization for improved cross-domain generalization.
citing papers explorer
-
Parameter-Efficient Conditioning for Material Generalization in Graph-Based Simulators
FiLM conditioning targeted at early message-passing layers lets pretrained GNS models generalize to new material properties using only 12 trajectories, a 5-fold data reduction versus multi-task baselines.
-
Gradual Domain Adaptation for Graph Learning
GGDA framework generates knowledge-preserving intermediate graphs via FGW metric and a vertex-based progression to enable gradual domain adaptation across large graph distribution shifts.
-
DIB-OD: Preserving the Invariant Core for Robust Heterogeneous Graph Adaptation via Decoupled Information Bottleneck and Online Distillation
DIB-OD isolates a stable invariant core in heterogeneous graph representations via orthogonal subspace decomposition, IB teacher-student distillation, HSIC independence, and confidence-gated regularization for improved cross-domain generalization.