GABE: Groupwise Affine Basis Encoding
Abstract
Introduce GABE (Groupwise Affine Basis Encoding) - a decomposition method that represents neural network weights as an addressable memory system. For any group of similar layers, we extract three interpretable components: a shared mean weight (long-term memory), a low-rank basis of variations (address space), and per-layer coefficients (pointers).
Fields of Study & Subjects
Explainable Artificial Intelligence (XAI)
Advanced Neural Network Applications
Neural Networks and Applications
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 5 February 2026
- Language
- EN