Skip to content
View d-corsi's full-sized avatar

Block or report d-corsi

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
d-corsi/README.md

Hi there 👋

I'm a Postdoctoral Researcher at University of California: Irvine, working in the Intelligent Dynamics Lab led by Prof. Roy Fox. Before that, I spent time as a visiting researcher with Prof. Guy Katz at the Hebrew University of Jerusalem and completed my PhD at the University of Verona under Prof. Alessandro Farinelli.

Research Interests 🔭

I’m passionate about advancing Deep Reinforcement Learning (DRL) for robotics, with a strong focus on making it safe and reliable in real-world, safety-critical environments. My research explores:

  • Safety in RL: constrained RL, formal verification, and shielding to ensure robots behave safely when deployed. Foundation models for robotics: as part of the DARPA SAFRON project, I am investigating how LLMs, VLMs, and VLAMs can be integrated into robotic control, with a special emphasis on safety — moving beyond treating these models as black boxes.
  • Generative & causal world models: building generative models that enable robots to predict the future and reason about consequences, while structuring these models with causal inference to improve their generalization across tasks and environments.

⚡ Bridging the gap between theory and practical robotics is at the heart of what I do ⚡

📫 Contact 📫

Pinned Loading

  1. eProVe eProVe Public

    This is the python implementation of ϵ-ProVe, a tool for the "AllDNN-Verification Problem" (i.e., the problem of computing the set of all the areas that do not result in a violation for a given DNN…

    Python

  2. lambda-PPO lambda-PPO Public

    Implementation of the lambda-PPO algorithm, an improved version of the standard Lagrangian Proximal Policy Optimization.

    Python 1

  3. NetworkVerifier NetworkVerifier Public

    A set of algorithms for the formal verification and analysis of Neural Networks, implemented in Python.

    Python 3 2

  4. BasicRL BasicRL Public

    A basic Tensorflow 2 implementation of the standard reinforcement learning algorithms, designed to solve gym-like environments.

    Python 4