AI safety researcher

Baochun Li

University of Toronto

Google Scholar dblp ORCID Homepage

Latest AI safety papers

On this site: 7 papers

June 2025

Leaner Training, Lower Leakage: Revisiting Memorization in LLM Fine-Tuning with LoRA
Jun 25arXiv ↗
Hear No Evil: Detecting Gradient Leakage by Malicious Servers in Federated Learning
Jun 25arXiv ↗

February 2024

FedReview: Review and Dispose Poisoned Updates without Validation Datasets or Historic Knowledge
Feb 26arXiv ↗

October 2022

Feature Reconstruction Attacks and Countermeasures of DNN training in Vertical Federated Learning
Oct 13arXiv ↗

January 2022

Towards Private Learning on Decentralized Graphs with Local Differential Privacy
Jan 23arXiv ↗

Authorship follows dblp and the author's own Google Scholar profile; only AI safety papers on this site are listed.