Lorenzo Bisi

Papers

Total Citations

H-Index

About

Lorenzo Bisi is a researcher advancing the frontiers of safe and reliable reinforcement learning, with a primary focus on risk-averse decision-making. His work addresses a critical challenge in AI: how to train agents that not only maximize rewards but also minimize exposure to catastrophic failures. Bisi’s major contributions center on bridging risk-averse and risk-neutral policy optimization, demonstrating that complex risk-sensitive objectives can be effectively tackled using standard, well-understood algorithms. His 2022 paper, "Risk-averse policy optimization via risk-neutral policy optimization," which has garnered 9 citations, introduces a transformative framework that simplifies the optimization of risk-averse policies, making them more accessible for real-world applications like finance and robotics. Building on this, his 2023 work on "Risk-averse optimization of reward-based coherent risk measures" further refines these techniques, offering a principled approach to handling coherent risk measures. Though early in his career, Bisi’s research is already shaping how practitioners design robust AI systems, earning recognition for its theoretical clarity and practical utility. His work stands as a vital resource for students and researchers seeking to build agents that are not only effective but also trustworthy in high-stakes environments.