Phuoc Nguyen

SEP 8, 2026 · PREPRINT

Geometry-Aware Bayesian Parameter-Efficient Fine-Tuning on the Stiefel Manifold via Stein Variational Gradient Descent

arXiv

This is an early-stage methodological paper presenting a novel computational framework for fine-tuning large models, lacking clinical or real-world validation and reporting only computational benchmarks without comparison to established clinical standards.