Showing cs.LGShow all
2 papers · 1 filter
cs.LG2026
PhysSAE: Mechanistic Interpretability of PINNs with Sparse Autoencoders
Nandita N. Patil, Eshwar R. A., Gajanan V. Honnavar
Physics-Informed Neural Networks (PINNs) embed PDE residuals into neural network training, but their internal representations remain opaque: it is unknown what physical features th…
cs.LG2026
Action-Inspired Generative Models
Eshwar R. A., Debnath Pal
We introduce Action-Inspired Generative Models (AGMs), a dual-network generative framework motivated by the observation that existing bridge-matching methods assign uniform regress…