Model-intrinsic, Weight-level, and Training-free Spectral Fingerprinting for Large Language Models

Summary

The paper proposes a training-free fingerprinting method for LLMs that works on weights alone. It builds spectral scalar features from the singular-value distributions of GATE projection weights in gated feed-forward networks, pools them by depth into a model-level fingerprint, and compares fingerprints with linear centered kernel alignment. Experiments on representative LLM families report clear separation between related and unrelated models and stable similarity under instruction tuning, quantization, pruning and distillation.