Event-grounded SAE analysis in VLA policies produces stronger causal effects on robot behavior than standard methods by anchoring features to clustered end-effector keyframes across simulations and real-robot tests.
Scaling and evaluating sparse autoencoders
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2representative citing papers
Aligned training reparameterizes SAEs to enforce unit alignment between encoder and decoder directions, yielding Pareto gains on SAEBench while removing dead features and improving stability.
citing papers explorer
-
Event-Grounded Sparse Autoencoders for Vision-Language-Action Policies
Event-grounded SAE analysis in VLA policies produces stronger causal effects on robot behavior than standard methods by anchoring features to clustered end-effector keyframes across simulations and real-robot tests.
-
Aligned Training: A Parameter-Free Method to Improve Feature Quality and Stability of Sparse Autoencoders (SAE)
Aligned training reparameterizes SAEs to enforce unit alignment between encoder and decoder directions, yielding Pareto gains on SAEBench while removing dead features and improving stability.