Vishalvermalabs (auto-discovered)

Hyperbolic Sparse Autoencoders (HyperSAE) outperform standard Euclidean Sparse Autoencoders (FlatSAEs) in reconstructing Google Gemma-2-2B activations, reducing reconstruction mean squared error (MSE)…

A 5-condition ablation study (3,175 total runs) validates that growth-ratio normalized Lyapunov energy functions serve as precise leading indicators of agent task failure, achieving zero stability vio…

Researchers propose Hyperbolic Sparse Autoencoders (HypSAEs) to overcome geometric limitations in mechanistic interpretability of large language models. By applying hyperbolic geometry as a weight-spa…