Davide Corsi

Thank you @yudapearl ! It's such an honor for me to co-organize the First Causal Reinforcement Learning Workshop at RLC 2025. We have a wonderful list of speakers from both the causality side and reinforcement learning side. Checkout our website here! sites.google.com/uci.edu/crlw20…

We’re honored to have Turing Award winner @yudapearl supporting our Causal Reinforcement Learning Workshop! Excited to see you at @RL_Conference! Check out our Call for Papers and the list of our amazing speakers: sites.google.com/uci.edu/crlw20…

Thrilled to announce that our paper has won one of the ‘Outstanding Paper Awards’ at #RLC! A huge thank you to my coauthors and the conference organizers for this honor @RL_Conference.

Looking for a challenging and realistic benchmark for your #DRL algorithm? Check out our new paper accepted at #RLC2024! Thanks to my co-authors @alex_sandrof and @dav_camponogara! Paper: arxiv.org/abs/2405.20534 Github: tinyurl.com/4jd773kb

How to face #safety challenges in DRL and #robotics? I am very excited to share that my first work in Prof. @roydfox's group at #UCI has been accepted at RLC2024! @RL_Conference Verification-Guided Shielding for Deep Reinforcement Learning: arxiv.org/abs/2406.06507

Our new paper has been accepted at #AAAI24🎉 A novel approach to identify the (un)safe regions in the domain of neural networks. "Enumerating Safe Regions in Deep Neural Networks with Provable Probabilistic Guarantees" Paper: arxiv.org/pdf/2308.09842… Git: github.com/d-corsi/eProVe

Interested in the intersection between formal verification and explainability in deep reinforcement learning? Check out our paper accepted at #FMCAD2023: "Formally Explaining Neural Networks within Reactive Systems"! 🤩 arxiv: arxiv.org/abs/2308.00143

Excited to share that our work "Constrained Reinforcement Learning and Formal Verification for Safe Colonoscopy Navigation" has been accepted to #IROS2023 🤖 Many thanks to my co-authors @lmarza_ and @its_pores! 🎉 arxiv: arxiv.org/abs/2303.03207

Our paper titled "The #'DNN-Verification problem: Counting Unsafe Inputs for Deep Neural Networks" has been accepted to IJCAI 2023 #IJCAI2023. Thanks to my co-author @lmarza_ 👏👏👏 arxiv: arxiv.org/abs/2301.07068