DynaDiff uses weight-graph diffusion with a functional consistency loss and dynamics-informed prompting to generate adapted predictors, reporting 10.78% average accuracy gains over baselines while amortizing adaptation cost offline.
Compared to these works, we innovatively treat the complete model weights as generated objects and explicitly model their joint distribution with the environment
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CE 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Generative Adaptation of Dynamics to Environmental Shifts via Weight-space Diffusion
DynaDiff uses weight-graph diffusion with a functional consistency loss and dynamics-informed prompting to generate adapted predictors, reporting 10.78% average accuracy gains over baselines while amortizing adaptation cost offline.