aliuyar1234/surgical-posttraining-diffing: Surgical Post-Training Diffing v0.1.0
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Initial public release of Surgical Post-Training Diffing. This release includes: source code for sparse PT-to-IT diffing, intervention masks, and evaluation experiment configs and test suite manuscript source in LaTeX final paper PDF Highlights: sparse answer-phase surrogate learning for PT-to-IT behavior shifts feature-mask intervention framework for capability and verbosity analysis evaluation code for fidelity, capability, verbosity, baselines, ablations, and bootstrap comparisons public-facing config cleanup with support for Hugging Face model IDs and relative local paths Included paper: paper/surgical-posttraining-diffing-ali-uyar.pdf Notes: this is a slim public source release large generated artifacts, caches, checkpoints, and internal planning documents are intentionally excluded some experiment configs reference artifacts that must be regenerated locally Author: Ali Uyar
Fields of Study & Subjects
Surgical Simulation and Training
Domain Adaptation and Few-Shot Learning
Topic Modeling
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 31 March 2026
- Language
- EN