TY - RPRT TI - Verifiable Fine-Tuning for LLMs: Zero-Knowledge Training Proofs Bound to Data Provenance and Policy AU - Hasan Akgul AU - Daniel Borg AU - Arta Berisha AU - Amina Rahimova AU - Andrej Novak AU - Mila Petrov PY - 2025 UR - https://arxiv.org/abs/2510.16830 ID - 2510.16830 ER -