errew/Statelens: StateLens v1.0.1: The Transformer Expansion System
Abstract
StateLens: The Transformer Expansion System Overview This release contains the experimental data and analysis code for our comprehensive study of Transformer internal dynamics. Key Findings Transformer layers are expansive systems: Layer 2 λ ≈ 1.7-1.8 > 1 K-θ Monotonicity Law: Validated across 15 models Temperature Modulation: τ = τ_min + C/β SI Classification: Continuous spectrum (LP → SP → OSC) Papers 📄 English Version 📄 中文版 Contents 📊 Data: validation_results/*.json 📈 Figures: validation_results/figures/ 🔧 Scripts: scripts/ Citation @dataset{statelens_expansion_2026, title={The Transformer Expansion System: Geometry of Representation and Dynamics of Mixing}, author={StateLens Project}, year={2026}, publisher={Zenodo}, doi={10.5281/zenodo.xxxxx} } License MIT License
Fields of Study & Subjects
Topic Modeling
Neural Networks and Applications
Advanced Neural Network Applications
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 11 March 2026
- Language
- EN