Semiotic Reflexive Transformer Adapters
28 repos
Specialized neural network adapter implementations for the Semiotic Reflexive Transformer (SRT) architecture, focusing on model interpretation and feature extraction across various model scales and variants. The cluster centers on adapter-based approaches for understanding and introspecting transformer model behavior, with central repositories representing different SRT model versions and configurations. While primarily non-Python repositories dominate the cluster, the shared focus on interpretability and reflexive introspection of transformer internals provides strong thematic coherence around making large language models more transparent and analyzable.
Python · 2
interpretability ·16
srt ·15
semiotic-reflexive-transformer ·13
introspection ·11
gradio ·10
frozen-backbone ·8
feature-extraction ·8
llm ·7
en ·6
custom_code ·6
