GFM applies conditional flow matching to neural weight trajectories, predicting final weights from a short observed prefix with accuracy competitive to Transformers on synthetic and CIFAR-10 tasks.
For reproducibility, we generate 50 trajectories per seed across 5 random seeds (0-4), where the first 30 trajectories use the 3-layer MLP and the remaining 20 use the 2-layer MLP
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Gradient Flow Matching for Learning Update Dynamics in Neural Network Training
GFM applies conditional flow matching to neural weight trajectories, predicting final weights from a short observed prefix with accuracy competitive to Transformers on synthetic and CIFAR-10 tasks.