Alessio Lomuscio
Professor of Safe Artificial Intelligence
Develops scalable verification methods and formal guarantees for neural networks, autonomous agents and neuro-symbolic systems.
Alessio Lomuscio's lab at Imperial works on reasoning & neuro-symbolic AI, embodied & active vision and compilers & programming languages. 34 papers since 2023. 7 PhD students. 12 alumni. Applications close .
At a glance
- papers, 2023–2026
- 9/ year
- at T1 venues
- 0%
- PhD students
- 7
- MS students
- 0
- alumni listed
- 12
Papers by yearfrom the lab homepage, 2023–2026
- 12 papers2026 · T1 0 · collecting
- 9 papers2025 · T1 0
- 4 papers2024 · T1 0
- 9 papers2023 · T1 0
T1 venuesOther
Research focuspaper titles that use each topic's words
- Reasoning & neuro-symbolic AI2 papers
- Compilers & programming languages2 papers
- Embodied & active vision0 papers
- develops scalable0 papers
- autonomous agents0 papers
Paper list
From the lab homepage.
202612 papers
- Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR26). Denver, CO, USA. Computer Vision Foundation / IEEE.A Robust Out-of-Distribution Detection Framework via Synergistic Smoothing (opens in a new tab)
- Proceedings of the AAAI26 Workshop on Artificial Intelligence for Air Transportation (AI4AT). Singapore.Certified Background-Invariant Training via Latent Generative Models
- Proceedings of the International Conference on Language Modeling (COLM26). San Francisco, CA, USA.Eliciting Intrinsic Hallucinations in LLMs via Semantically Equivalent Adversarial Attacks (opens in a new tab)
- Forthcoming, NeurIPSEnd-to-End Verification of Neuro-symbolic Automata via Contrastive Logit-Gaps
- Proceedings of the International Symposium on AI Verification (SAIV26). Lisbon, Portugal.Hybrid Robustness Verification for Spatio-temporal Neural Networks (opens in a new tab)
- Proceedings of the AAAI26 Workshop on Artificial Intelligence for Cyber Security (AICS). Singapore.Improving Neural Network Robustness to Convolutional Perturbations Through Certified Training (opens in a new tab)
- NeurIPS26 Workshop on Grounded User Simulation for Model Evaluation and Training: Diversity, Fidelity, and Validity (UserSim).Intent-Preserving Adversarial User Simulation for Multi-Turn Agent Evaluation (opens in a new tab)
- Proceedings of the European Conference on Computer Vision (ECCV26). Malmoe, Sweden. Springer.IoUCert: Robustness Verification for anchor-based Object Detectors (opens in a new tab)
- Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR26). Denver, CO, USA. Computer Vision Foundation / IEEE.Lipschitz Optimization for Formal Verification of Homographies (opens in a new tab)
- Proceedings of the IJCAI26 Workshop on Explainable AI. Bremen, Germany.Out-of-Distribution Detection using Counterfactual Distance (opens in a new tab)
- Proceedings of the International Conference on Autonomous Agents and Multi-agent Systems (AAMAS26). Paphos, Cyprus. IFAAMAS Press.Safe and Scalable Multi-agent Coordination with Reconstructed Level-k Monte Carlo Tree Search (opens in a new tab)
- NeurIPS26 Workshop Verify-Agents. Poster.Verifying the Verifier: What a Guardrail Certificate Pays For (opens in a new tab)
20259 papers
- Proceedings of the International Conference on Neurosymbolic Reasoning (NeSY25). Santa Cruz, CA, USA. (Outstanding paper award).A Scalable Approach to Probabilistic Neuro-symbolic Robustness Verification (opens in a new tab)
- Proceedings of the AAAI Conference on Artificial Intelligence (AAAI25). Philadelphia, PA, USA. AAAI Press.Dynamic Back-substitution in Bound-propagation-based Neural Network Verification (opens in a new tab)
- Proceedings of the CVPR2025 Workshop on Safe Artificial Intelligence for All Domains (SAID). Nashville, TN, USA.Improving Weather-based OOD Generalisation in Lidar-based Perception Models via Adversarial Training (opens in a new tab)
- Proceedings of the Conference on Uncertainty in Artificial Intelligence (UAI25). Rio de Janeiro, Brazil.Learning Robust XGBoost Ensembles for Regression Tasks (opens in a new tab)
- Proceedings of the International Conference on Autonomous Agents and Multi-agent Systems (AAMAS25). Detroit, MI, USA. IFAAMAS Press.LTL Verification of Memoryful Neural Agents (opens in a new tab)
- Proceedings of the IEEE/CVF Winter Workshops of the Conference on Applications of Computer Vision (WACV25). Tucson, AZ, USA. CVF Press.Robustness to Perturbations in the Frequency Domain: Neural Network Verification and Certified Training (opens in a new tab)
- Proceedings of the Annual Conference on Neural Information Processing Systems (NeurIPS25). San Diego, CA, USA.Scalable Neural Network Geometric Robustness Validation via Hölder Optimisation (opens in a new tab)
- Proceedings of the European Conference on Artificial Intelligence (ECAI25). Bologna, Italy.Verifiably Robust Contrastive Learning (opens in a new tab)
- Proceedings of the AAAI Conference on Artificial Intelligence (AAAI25). Philadelphia, PA, USA. AAAI Press.Verification of Neural Networks against Convolutional Perturbations via Parameterised Kernels (opens in a new tab)
20244 papers
- Proceedings of the International Conference on Learning Representations (ICLR24). Vienna, Austria.Expressive Losses for Verified Robustness via Convex Combinations (opens in a new tab)
- Proceedings of the ACM International Conference on AI in Finance (ICAIF24). Brooklyn, New York, NY, USA.Reducing Return Volatility in Neural Network-Based Asset Allocation via Formal Verification and Certified Training (opens in a new tab)
- Proceedings of the International Conference on Artificial Intelligence and Statistic (AISTATS24). Valencia, Spain.Tight Verification of Probabilistic Robustness in Bayesian Neural Networks (opens in a new tab)
- Proceedings of the European Conference on Artificial Intelligence (ECAI24). Santiago de Compostela, Spain.Verification of Geometric Robustness of Neural Networks via Piecewise Linear Approximation and Lipschitz Optimisation (opens in a new tab)
20239 papers
- Proceedings of the AAAI Conference on Artificial Intelligence (AAAI23). Washington, DC, USA. Journal Track.A Counter Abstraction Technique for Verifying Properties of Probabilistic Swarm Systems (opens in a new tab)
- Proceedings of the AAAI Conference on Artificial Intelligence (AAAI23). Washington, DC, USA.A Semidefinite Relaxation based Branch-and-Bound Method for Tight Neural Network Verification (opens in a new tab)
- Proceedings of the International Conference on Autonomous Agents and Multiagent Systems (AAMAS23). London, UK. IFAAMAS Press.Bounded and Unbounded Verification of RNN-Based Agents in Non-deterministic Environments (extended abstract) (opens in a new tab)
- Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR23). Montreal, Canada. Computer Vision Foundation / IEEE.Efficient Verification of Neural Networks against LVM-based Specifications (opens in a new tab)
- Proceedings of the AAAI Conference on Artificial Intelligence (AAAI23). Washington, DC, USA.Iteratively Enhanced Semidefinite Relaxations for Efficient Neural Network Verification (opens in a new tab)
- Proceedings of the AAAI Conference on Artificial Intelligence (AAAI23). Washington, DC, USA.Robust Training of Neural Networks against Bias Field Perturbations (opens in a new tab)
- Proceedings of the International Conference on Autonomous Agents and Multiagent Systems (AAMAS23). London, UK. IFAAMAS Press.Towards Robust Contrastive Explanations for Human-Neural Multi-Agent Systems (extended abstract) (opens in a new tab)
- Proceedings of the International Conference on Principles of Knowledge Representation and Reasoning (KR23). Rhodes, Greece. AAAI Press.Verification of Semantic Key Point Detection for Aircraft Pose Estimation (opens in a new tab)
- Proceedings of the International Conference on Neural Networks (IJCNN23). Gold Coast, Australia. IEEE Press.Verification-friendly Networks: the Case for Parametric ReLUs (opens in a new tab)
Where alumni went12 listed
Apply to Imperial2027 (Apr/Oct)
Computing PhD
- Round
- Round 2 (1 Dec 2026)
- Opens
- —
- GRE
- See program page
- English
- See program page
- Fee
- none
Confirmed on the program page.
Similar labsnearest by research summary
- USCMU
Artur W. Dubrawski
- Recognition, detection & segmentation
- USStanford
James Zou
- Trustworthy & safe ML
- CNTsinghua
Guiguang Ding
- Efficient training & inference
- SGNUS
Bryan Hooi
- Time series, anomaly & spatio-temporal data
- USUT Austin
Haris Vikalo
- Trustworthy & safe ML
- CNPeking
Xiaotie Deng
- Reasoning & neuro-symbolic AI
Your notes
Save this lab to keep notes.
Data
Some numbers could not be verified:
- The lab site and the paper record do not agree.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.
Imperialcloses