Sparse autoencoders trained on frozen MeshGraphNets embeddings yield feature dictionaries whose mesh-space activations align with high-vorticity regions better than embedding-norm, PCA, or random baselines.
Comparative study of machine learning tech- niques for post-combustion carbon capture systems
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CE 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Interpreting CFD Surrogates through Sparse Autoencoders
Sparse autoencoders trained on frozen MeshGraphNets embeddings yield feature dictionaries whose mesh-space activations align with high-vorticity regions better than embedding-norm, PCA, or random baselines.