Seonglae ChoDepends on seed and dataset (SAE Training)
Weight Cosine Similarity
Orphan features still shows high interpretability which indicates the different seed may have found the subset of the “idealized dictionary size”.
- seed - weight initialization matters → SAE weight initialization helps to prevent this issue
- Weight cosine similarity + Hungarian Matching
It sets 1 - cosine similarity matrix to cost matrix and applies Hungarian Matching to find optimal 1:1 matching
- dataset - matters more than seed
FaithfulSAE FaithfulSAEseonglae • Updated 2025 Aug 21 15:27
FaithfulSAE
seonglae • Updated 2025 Aug 21 15:27
High-frequency features and feature set stability are problems in SAE. This approach attempts to solve these issues by eliminating external dataset dependency and increasing faithfulness through self-synthetic datasets. It defines high-frequent features as FFR (Fake Feature Ratio) and uses the existing SFR (Shared Feature Ratio) to demonstrate that Faithful SAE improves both metrics.
This is a rare attempt to improve SAE from a dataset perspective, highlighting the issue of external dependencies in interpretability.