Answers are grounded in this paper and cite their sources.
A framework analyzes post-training modifications in pretrained models, indicating new techniques for AI system development.
Model Sculpting presents a unified framework for describing and analyzing post-training modifications of pretrained models. It formalizes interventions as structured operators acting along structural, parametric, functional, and behavioral dimensions, enabling systematic comparison and composition of techniques such as fine-tuning, editing, alignment, compression, and routing. The framework introduces formal definitions, distance measures, and theoretical properties to support consistent analysis, benchmarking, and model-centric development of AI systems.
No takes yet. Share an insight, caveat, or question.
Rayyan Khan (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: