Scholar

Hengrui Jia

Google Scholar ID: g2vBgnoAAAAJ

University of Toronto and Vector Institute

Deep LearningAdversarial Machine Learning

Citations & Impact

All-time

Citations

2,275

H-index

i10-index

Publications

Co-authors

list available

Contact

Publications

2 items

2025

Cited

2025

Cited

Resume (English only)

Academic Achievements

Publications:
- Backdoor Detection through Replicated Execution of Outsourced Training
- LLM Dataset Inference: Did you train on my dataset?
- Gradients Look Alike: Sensitivity is Often Overestimated in DP-SGD
- Proof-of-Learning is Currently More Broken Than You Think
- On the Necessity of Auditable Algorithmic Definitions for Machine Unlearning
- A Zest of LIME: Towards Architecture-Independent Model Distances
- SoK: Machine Learning Governance
- Proof-of-Learning: Definitions and Practice
- Entangled Watermarks as a Defense against Model Extraction
- Machine Unlearning
Awards:
- Ontario Graduate Scholarship (2024-2025)
- Ontario Graduate Scholarship (2023-2024)
- Mary H. Beatty Fellowship (2022-2023)
- Vector Scholarship in Artificial Intelligence (2020-2021)
- Dean’s List (2016-2020)
Invited Talks:
- Ownership Resolution in ML, Purdue University (2024)
- Ownership Resolution in ML, Northwestern University (2024)
- Ownership Resolution in ML, University of Wisconsin–Madison (2024)
- Ownership of ML Models, Mila - Quebec AI Institute (2024)
- Entangled Watermarks as a Defense against Model Extraction, DeepMind (2023)
- A Zest of LIME: Towards Architecture-Independent Model Distances, Workshop on Algorithmic Audits of Algorithms (2023)
- Entangled Watermarks as a Defense against Model Extraction, Intel (2022)
- Machine Unlearning, Vector Institute (2021)

Research Experience

Education

PhD student at the University of Toronto and Vector Institute, advised by Prof. Nicolas Papernot.

Background

Research interests: the intersection of security and machine learning, or trustworthy machine learning. Particularly interested in answering questions such as: what risks may come along with the benefits brought by machine learning, who is responsible for such risks, and how can we mitigate them?

Miscellany