Jinfeng Yi

About

Jinfeng Yi is a leading researcher in the security and privacy of machine learning, with a particular focus on deep reinforcement learning (DRL). His major contributions center on exposing critical vulnerabilities in DRL systems, demonstrating how adversarial attacks can manipulate these models by adding small perturbations to observations, even when the attacker has limited access to the victim model. Yi’s work also pioneers the study of privacy-leakage attacks, revealing that an agent’s behavior can inadvertently expose sensitive training data—a novel threat that extends beyond traditional data breaches in natural language or facial recognition. His most-cited paper, “Characterizing Attacks on Deep Reinforcement Learning” (2019, 52 citations), provides a foundational framework for understanding attack vectors, while his privacy-focused studies (totaling over 50 citations) highlight the dual risks of manipulation and exposure in autonomous systems. By bridging adversarial robustness and privacy preservation, Yi’s research has profound implications for deploying DRL in safety-critical domains like autonomous driving and healthcare. His work is essential reading for anyone studying the intersection of AI security, reinforcement learning, and trustworthy machine learning.

Research Focus

Key Achievements

Top Papers

1. 1
   Characterizing Attacks on Deep Reinforcement Learning

   52 citations · 2019
2. 2
   How You Act Tells a Lot: Privacy-Leaking Attack on Deep Reinforcement Learning

   38 citations · 2019
3. 3
   How You Act Tells a Lot: Privacy-Leakage Attack on Deep Reinforcement Learning

   14 citations · 2019

Key Collaborators

Contact & Links