About this resource
<p class="">This repository contains the complete experimental framework and raw data results required to replicate the findings in the manuscript "Scale-Free Attention Dynamics in Transformer-Based Language Models."</p> <p class=""><strong>Contents:</strong></p> <ul> <li><strong>Code:</strong> Jupyter notebooks for extracting attention weight matrices from multiple transformer architectures (Llama-3, GPT-2, BERT, T5, Pythia) and conducting rigorous Clauset-Shalizi-Newman (CSN) power-law fitting. It includes control experiments (Zipfian vs. Uniform training) and ablation studies (attention sink masking).</li> <li><strong>Results:</strong> Raw CSV outputs containing the calculated power-law scaling exponents (alpha), Kolmogorov-Smirnov statistics, and Vuong likelihood ratio test verdicts across all six experiments.</li> </ul>
Files and access
Resources are read online through DARE. Direct file links are not published; access is checked against DSpace when a document is opened. Authorised librarian workflows are unaffected.
This record contains bibliographic metadata, but no files are currently attached.
