Ren, Shaolei
20 publications
ICLRW
2025
ProDiF: Protecting Domain-Invariant Features to Secure Pre-Trained Models Against Extraction
ICLR
2024
ArchLock: Locking DNN Transferability at the Architecture Level with a Zero-Cost Binary Predictor
NeurIPS
2024
Bileve: Securing Text Provenance in Large Language Models Against Spoofing with Bi-Level Signature
NeurIPS
2023
Beyond Black-Box Advice: Learning-Augmented Algorithms for MDPs with Q-Value Predictions