Hamed Ayoobi

London, UK and Amsterdam, NL · h.ayoobi@imperial.ac.uk

Postdoctoral Research Associate at Imperial College London with deep expertise in Generative AI (LLMs, VLMs, LMMs), Explainable AI (XAI), Computer Vision (Medical Imaging & Robotic Vision), and Retrieval-Augmented Generation (RAG). Proven experience deploying containerized AI systems on AWS/Azure, developing explainable ML frameworks, and supervising/leading multidisciplinary research and engineering teams.

My work aims to make AI systems more transparent, understandable, and trustworthy, particularly through prototypical and argumentative approaches. I am passionate about bridging the gap between complex AI models and human comprehension, contributing to safer and more reliable AI deployments.

Research Experience

Postdoctoral Research Associate

Imperial College London, UK

Advanced research in Explainable AI (XAI) and Responsible AI for interpreting Deep Neural Networks (DNNs), including LLMs, LVMs, CNNs, and Large Multimodal Models for different applications including Medical Image Analysis, Robotic Vision, RAG, and Agent-Based AI systems.

Key Projects:
  • SpArX: Sparse Argumentative Explanations for DNNs.
  • ProtoArgNet: Interpretable image classification using prototypes and argumentation.
  • LLM-PAI: Large Language Models and Prototypical Argumentative Interpretability.
  • CAFE: Conflict-Aware FEature attribution for multi-modal networks.
  • ProtoArgDebug: Automated debugging for DNNs based on expert feedback.
  • A comprehensive survey on interpretable prototypical learning in LLMs, VLMs, and CNNs.
  • Collaborating with National Health Service (NHS in the UK) to apply XAI methods to data from 50,000+ adults diagnosed with a primary brain tumour.
Mar 2022 - Present

Research Intern

ABN AMRO Bank, Amsterdam, Netherlands

Building Retrieval-Augmented Generation (RAG) systems and Agent-based LLMs for customer service chatbots using various Transformers (BERT, RoBERTa, GPT, LLaMA, T5, etc.). Utilized AWS SageMaker, Azure ML/DataBricks/Cognitive Services, and vector databases like ChromaDB, FAISS, Azure Cognitive Search, and graph vector databases (Neo4j, Cypher).

Jun 2021 - Aug 2021

Research Intern

Linnaeus University, Vaxjo, Sweden

Researching on a project related to online incremental machine learning and deep neural networks.

Feb 2017 - Aug 2017

Education

University of Groningen, Netherlands

PhD in Artificial Intelligence
  • Proposing an explainable online incremental machine learning technique called Argumentation-Based Learning (ABL) using argumentation theory.
  • Proposing an open-ended 3D object recognition technique called Local Hierarchical Dirichlet Process for the robotic manipulation task.
  • Proposing an explainable machine learning technique for 3D object recognition and segmentation.
Mar 2018 - Feb 2022

Yazd University, Yazd, IR

Master in Artificial Intelligence & Robotics

Graduated with honors (cum laude), GPA 18.56/20

Sep 2016 - Feb 2018

Publications

For a comprehensive and up-to-date list of my publications, please visit my Google Scholar profile.

Selected Conference Proceedings
ProtoArgNet: Interpretable Image Classification with Super-Prototypes and Argumentation
Hamed Ayoobi, Nico Potyka, Francesca Toni
AAAI 2025 - 39th Annual AAAI Conference on Artificial Intelligence, Philadelphia, Pennsylvania, USA
Hidden Conflicts in Neural Networks and their Implications for Explainability
Adam Dejl, Dekai Zhang, Hamed Ayoobi, Matthew Williams, Francesca Toni
FAccT '25: The 2025 ACM Conference on Fairness, Accountability, and Transparency Proceedings, Athens, Greece
Contestable AI needs Computational Argumentation
Francesco Leofante, Hamed Ayoobi, Adam Dejl, Gabriel Freedman, Deniz Gorur, Junqi Jiang, Guilherme Paulino-Passos, Antonio Rago, Anna Rapberger, Fabrizio Russo, Xiang Yin, Dekai Zhang, Francesca Toni
KR 2024 - 21st International Conference on Principles of Knowledge Representation and Reasoning, Hanoi, Vietnam
Explain What You See: Open-Ended Segmentation and Recognition of Occluded 3D Objects
Hamed Ayoobi, Hamidreza Kasaei, Ming Cao, Rineke Verbrugge, Bart Verheij
ICRA 2023 - IEEE International Conference on Robotics and Automation, London, United Kingdom
SpArX: Sparse Argumentative Explanations for Neural Networks
Hamed Ayoobi, Nico Potyka, Francesca Toni
ECAI 2023 - 26th European Conference on Artificial Intelligence, Kraków, Poland
Argue to Learn: Accelerated Argumentation-Based Learning
Hamed Ayoobi, Ming Cao, Rineke Verbrugge, Bart Verheij
ICMLA 2021 - 20th IEEE International Conference on Machine Learning and Applications, Pasadena, CA, USA
Handling Unforeseen Failures Using Argumentation-Based Learning
Hamed Ayoobi, Ming Cao, Rineke Verbrugge, Bart Verheij
CASE 2019 - 15th IEEE International Conference on Automation Science and Engineering, Vancouver, BC, Canada
Journal Articles
Argumentation-Based Online Incremental Learning
Hamed Ayoobi, Ming Cao, Rineke Verbrugge, Bart Verheij
IEEE Transactions on Automation Science and Engineering (T-ASE) 19.4 (2022) pp. 3419-3433
Local-HDP: Interactive open-ended 3D object category recognition in real-time robotic scenarios
Hamed Ayoobi, S. Hamidreza Kasaei, Ming Cao, Rineke Verbrugge, Bart Verheij
Robotics and Autonomous Systems (RAS) 147 (2022) p. 103911
Swift Distance Transformed Belief Propagation using a Novel Dynamic Label Pruning Method
Hamed Ayoobi, Mehdi Rezaeian
IET Image Processing 14.9 (2020) pp. 1822-1831
Preprints
Argumentative Debate for Transparent Bias Detection in AI
Hamed Ayoobi, Nico Potyka, Anna Rapberger, Francesca Toni (2025)
A comprehensive survey on interpretable prototypical learning in LLMs, VLMs, and CNNs
Hamed Ayoobi, Nico Potyka, Francesca Toni (2025)
LLM-PAI: Large Language Models and Prototypical Argumentative Interpretability
Hamed Ayoobi, Nico Potyka, Francesca Toni (2025)
GraspCaps: A Capsule Network Approach for Familiar 6DoF Object Grasping
Tomas Velde, Hamed Ayoobi, Hamidreza Kasaei (2024)

Supervision

PhD Students

  • Co-supervision of Adam Dejl for explaining Large Language Models (LLMs) and multi-modal deep learning models, Imperial College London (2022-Now).

Master Students

  • Co-supervision of three master students on projects related to XAI and Deep Learning for LLMs, images and tabular data, Imperial College London (2024-2025) and University of Groningen, The Netherlands (2022-2023).
  • Co-supervision of one master student on a project related to interpretable image classification using prototypical-part learning approaches at Cardiff University, Wales, UK (2024).

Bachelor Students

  • Co-supervision of four bachelor students in projects related to Machine Learning, Robotic Vision and Argumentation, University of Groningen, The Netherlands (2021-2023).
  • Co-supervising a group of six software engineering bachelor students for their graduation project in developing an interactive Python library for generating argumentative explanations for neural networks leading to a demo paper entitled "PySpArX: A Python library for generating Sparse Argumentative explanations for neural networks" published in ICLP 2023, Imperial College London (2023).

Teaching

Guest Lecturer

  • Knowledge and Agent Systems (KAS) course for two semesters (2022-2023 and 2023-2024) at the University of Groningen, the Netherlands.
  • Cognitive Robotic course at the University of Groningen, the Netherlands.

Lecturer

  • Deep Neural Networks, Yazd University, 2018.

Teaching Assistant

  • Arguing Agents Course, University of Groningen, 2021-2022.
  • Cognitive Robotics Course, University of Groningen, 2020-2021 and 2021-2022.
  • Deep Neural Networks, Yazd University, 2017-2018.
  • Computer Vision, Yazd University, 2017-2018.

Talks and Outreach

  • Invited Speaker for the 30th anniversary of AI at the University of Groningen to talk about AI's past, present and future.

Skills

Technical Expertise
  • Explainable AI (XAI), Machine Learning (ML), Natural Language Processing (LLMs), Argumentation-Based Learning, Computer (Robotic) Vision (LVMs, CNN), Deep Learning
Programming Languages & Tools

Python (Pandas, NumPy), C#.NET, Java, C/C++, OpenCV, PCL, HTML/CSS, JavaScript, SQL, PySpark

Machine Learning Frameworks & Libraries

Pytorch, Tensorflow, GPT, RAG, LangChain, LangGraph, Generative AI, Diffusers, Hugging Face Transformers, LlamaIndex, TruLens, ChromaDB, FAISS, Azure Cognitive Services, Neo4j, Cypher, NLP, Keras, Scikit-learn, Scipy, NLTK, Gensim etc.

Cloud Platforms

AWS SageMaker, Azure ML, Azure Databricks, Azure Cognitive Services.

Miscellaneous Technical Skills

Scrum, Linux, LaTeX (Overleaf Markdown), Tableau, Microsoft Office, Git.

Soft Skills

Team Management, Team Player, Time Management, Problem-solving, Engaging Presentation, Flexibility, Adaptability.

Languages
  • English: Advanced Professional proficiency
  • Dutch: B1

Professional Service

PC Member and Reviewer

AAAI 2025 & 2024, ICRA 2025 & 2023, IJCAI 2025 & 2024, KR 2024 & 2023, ECAI 2025 & 2024 & 2023, AISTATS 2024, XAI 2024 & 2023, XLoKR 2024 & 2023, IROS 2023, XAI-FIN 2023 & 2022, RH12023, DAMI Journal, Scientific Reports - Nature, Supercomputing Journal.

Organizing Seminars
  • Organizer of XAI Seminars at Imperial College London.

Awards, Funding & Certificates

Funding and Grants

My research is funded by:

  • H2020 ERC Argumentation-based Deep Interactive Explanations.
  • J.P. Morgan / Royal Academy of Engineering Chair in Argumentation for Interactive Explainable AI.
  • Research Chairs and Senior Research Fellowships - Argumentation for interactive explainable AI.

Funding Applications (Under Preparation):

  • NWO-VENI 2025 pre-proposal and full proposal.
  • NWO-XS 2025 proposal.
Certificates (Coursera)
  • Building and Evaluating Advanced RAG (DeepLearning.AI)
  • Generative AI with Large Language Models (LLMs) (Coursera, DeepLearning.AI)
  • Explainable Artificial Intelligence (XAI) (Coursera)
  • Databricks Fundumentals (Databricks Academy)
  • Neural Networks and Deep Learning (Coursera)
  • Structuring Machine Learning Projects (Coursera)
  • Improving Deep Neural Networks: Hyper-parameter Tuning, Regularization and Optimization (Coursera)
  • SQL Data Management and Analysis (Coursera)
  • Fundamentals of Scalable Data Science (Coursera)

Latest News

Stay tuned for updates on my latest research, projects, and activities!