Luca Marzari
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Luca Marzari has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Alessandro Farinelli
University of Verona
4 shared publications - Ferdinando Cicalese
Department of Computer Science, University of Verona, Verona, Italy
3 shared publications - Enrico Marchesini
Massachusetts Institute of Technology, Cambridge, MA, USA
2 shared publications - Christopher Amato
Computer Science and Artificial Intelligence Laboratory, MIT, Cambridge, MA, USA
1 shared publication - Francesco Leofante
Imperial College London
1 shared publication - Francesco Trotti
Department of Computer Science, University of Verona, Verona, Italy
1 shared publication
Latest Publications
- 30 December 2025
- Published by AI Access Foundation in Journal of Artificial Intelligence Research
- Vol. 84
Probabilistically Tightened Linear Relaxation-based Perturbation Analysis for Neural Network Verification
We present Probabilistically Tightened Linear Relaxation-based Perturbation Analysis (PT-LiRPA), a novel framework that combines over-approximation techniques from LiRPA-based approaches with a sampling-based method to compute tight intermediate reachable sets. In detail, we show that with negligible computational overhead, PT-LiRPA exploiting the estimated reachable sets, significantly tightens the lower and upper linear bounds of a neural network's output, reducing the computational cost of formal verification tools while providing probabilistic guarantees on verification soundness. Extensive experiments on standard formal verification benchmarks, including the International Verification of Neural Networks Competition, show that our PT-LiRPA-based verifier improves robustness certificates, i.e., the certified lower bound of ε perturbation tolerated by the models, by up to 3.31X and 2.26X compared to related work. Importantly, our probabilistic approach results in a valuable solution for challenging competition entries where state-of-the-art formal verification methods fail, allowing us to provide answers with high confidence (i.e., at least 99%).
- 30 December 2025
- Published by AI Access Foundation in Journal of Artificial Intelligence Research
- Vol. 84
Probabilistically Tightened Linear Relaxation-based Perturbation Analysis for Neural Network Verification
We present Probabilistically Tightened Linear Relaxation-based Perturbation Analysis (PT-LiRPA), a novel framework that combines over-approximation techniques from LiRPA-based approaches with a sampling-based method to compute tight intermediate reachable sets. In detail, we show that with negligible computational overhead, PT-LiRPA exploiting the estimated reachable sets, significantly tightens the lower and upper linear bounds of a neural network's output, reducing the computational cost of formal verification tools while providing probabilistic guarantees on verification soundness. Extensive experiments on standard formal verification benchmarks, including the International Verification of Neural Networks Competition, show that our PT-LiRPA-based verifier improves robustness certificates, i.e., the certified lower bound of ε perturbation tolerated by the models, by up to 3.31X and 2.26X compared to related work. Importantly, our probabilistic approach results in a valuable solution for challenging competition entries where state-of-the-art formal verification methods fail, allowing us to provide answers with high confidence (i.e., at least 99%).
- Open Access
- 8 December 2025
- Published by Elsevier in Artificial Intelligence
- Vol. 351
Probabilistically robust counterfactual explanations under model changes
- 8 December 2025
- Published by Elsevier in Artificial Intelligence
- Vol. 351
Open AccessProbabilistically robust counterfactual explanations under model changes
- Open Access
- 18 November 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Intelligent Systems and Technology
- Vol. 17 (1) , 1-27
Verifying Online Safety Properties for Safe Deep Reinforcement Learning
Ensuring safety in reinforcement learning (RL) is critical for deploying agents in real-world applications. During training, current safe RL approaches often rely on indicator cost functions that provide sparse feedback, resulting in two key limitations: (i) poor sample efficiency due to the lack of safety information in neighboring states, and (ii) dependence on cost-value functions, leading to brittle convergence and suboptimal performance. After training, safety is guaranteed via formal verification (FV) methods for deep neural networks, whose computational complexity hinders their application during training. We address the limitations of using cost functions via verification by proposing a safe RL method based on a violation value—the risk associated with policy decisions in a portion of the state space. Our approach verifies safety properties (i.e., state-action pairs) that may lead to unsafe behavior, and quantifies the size of the state space where properties are violated. This violation value is then used to penalize the agent during training to encourage safer policy behavior. Given the NP-hard nature of FV, we propose an efficient, sample-based approximation with probabilistic guarantees to compute the violation value. Extensive experiments on standard benchmarks and real-world robotic navigation tasks show that violation-augmented approaches significantly improve safety by reducing the number of unsafe states encountered while achieving superior performance compared to existing methods.
- 18 November 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Intelligent Systems and Technology
- Vol. 17 (1) , 1-27
Open AccessVerifying Online Safety Properties for Safe Deep Reinforcement Learning
Ensuring safety in reinforcement learning (RL) is critical for deploying agents in real-world applications. During training, current safe RL approaches often rely on indicator cost functions that provide sparse feedback, resulting in two key limitations: (i) poor sample efficiency due to the lack of safety information in neighboring states, and (ii) dependence on cost-value functions, leading to brittle convergence and suboptimal performance. After training, safety is guaranteed via formal verification (FV) methods for deep neural networks, whose computational complexity hinders their application during training. We address the limitations of using cost functions via verification by proposing a safe RL method based on a violation value—the risk associated with policy decisions in a portion of the state space. Our approach verifies safety properties (i.e., state-action pairs) that may lead to unsafe behavior, and quantifies the size of the state space where properties are violated. This violation value is then used to penalize the agent during training to encourage safer policy behavior. Given the NP-hard nature of FV, we propose an efficient, sample-based approximation with probabilistic guarantees to compute the violation value. Extensive experiments on standard benchmarks and real-world robotic navigation tasks show that violation-augmented approaches significantly improve safety by reducing the number of unsafe states encountered while achieving superior performance compared to existing methods.
- 6 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 10 (10) , 9630-9637
Designing Control Barrier Function via Probabilistic Enumeration for Safe Reinforcement Learning Navigation
Achieving safe autonomous navigation systems is critical for deploying robots in dynamic and uncertain real-world environments. In this paper, we propose a hierarchical control framework leveraging neural network verification techniques to design control barrier functions (CBFs) and policy correction mechanisms that ensure safe reinforcement learning navigation policies. Our approach relies on probabilistic enumeration to identify unsafe regions of operation, which are then used to construct a safe CBF-based control layer applicable to arbitrary policies. We validate our framework both in simulation and on a real robot, using a standard mobile robot benchmark and a highly dynamic aquatic environmental monitoring task. These experiments demonstrate the ability of the proposed solution to correct unsafe actions while preserving efficient navigation behavior. Our results show the promise of developing hierarchical verification-based systems to enable safe and robust navigation behaviors in complex scenarios.
- 6 August 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 10 (10) , 9630-9637
Designing Control Barrier Function via Probabilistic Enumeration for Safe Reinforcement Learning Navigation
Achieving safe autonomous navigation systems is critical for deploying robots in dynamic and uncertain real-world environments. In this paper, we propose a hierarchical control framework leveraging neural network verification techniques to design control barrier functions (CBFs) and policy correction mechanisms that ensure safe reinforcement learning navigation policies. Our approach relies on probabilistic enumeration to identify unsafe regions of operation, which are then used to construct a safe CBF-based control layer applicable to arbitrary policies. We validate our framework both in simulation and on a real robot, using a standard mobile robot benchmark and a highly dynamic aquatic environmental monitoring task. These experiments demonstrate the ability of the proposed solution to correct unsafe actions while preserving efficient navigation behavior. Our results show the promise of developing hierarchical verification-based systems to enable safe and robust navigation behaviors in complex scenarios.