Welcome!

I am a Ph.D. candidate and research associate in the Collaborative Artificial Intelligence (CAI) group at the University of Stuttgart, supervised by Prof. Dr. Andreas Bulling. I am also a Ph.D. scholar with the International Max Planck Research School for Intelligent Systems (IMPRS-IS). My IMPRS-IS Thesis Advisory Committee includes Prof. Dr. Georg Martius (Max Planck Institute for Intelliegent Systems, University of Tübingen) and Dr. Charley Wu (University of Tübingen), alongside Prof. Bulling.

I research cooperative AI with a focus on social reasoning and generalisation. Specifically, I work on multi-agent reinforcement learning (MARL), ad-hoc teamwork, large language models (LLMs), and Theory of Mind (ToM) in cooperative tasks. My long-term goal is to train agents capable of ad-hoc teamwork with diverse partners in unknown, open-ended environments. Among other things I research ...
  1. ... the intersection between partner and environment generalisation. In The Overcooked Generalisation Challenge we introduced a benchmark where agents must cooperate with novel partners in unknown environments. In Unsupervised Partner Design (UPD), we proposed a method addressing this challenge.
  2. ... how computational Theory of Mind can be used to improve human-AI cooperation. In the Yokai Learning Environment we introduced a new benchmark for social reasoning, and in our work on LLMs we evaluated their Theory of Mind capabilities (e.g. 1, 2).
You can find more about me or my work on the about page, on our groups website, or on google scholar. For students interested in a project or thesis please check the open thesis page.

Publications

Published

  1. 2025: Bortoletto, M., Ruhdorfer, C. & Bulling, A. (2025). ToM-SSI: Evaluating Theory of Mind in Situated Social Interactions. Proc. Empirical Methods in Natural Language Processing (EMNLP) [website] [arxiv]
  2. 2025: Bortoletto, M., Ruhdorfer, C., Shi, L. & Bulling, A. (2025). Brittle Minds, Fixable Activations: Understanding Belief Representations in Language Models. Findings of Empirical Methods in Natural Language Processing (EMNLP Findings) [website] [arxiv]
  3. 2025: Ruhdorfer, C., Bortoletto, M., Penzkofer, A. & Bulling, A. (2025). The Overcooked Generalisation Challenge: Evaluating Cooperation with Novel Partners in Unknown Environments Using Unsupervised Environment Design. Transactions on Machine Learning Research. [paper] [website] [arxiv]
  4. 2025: Ruhdorfer, C., Bortoletto, M. & Bulling, A. (2025) The Yōkai Learning Environment: Tracking Beliefs Over Space and Time. IJCAI Workshop on Generative AI & Theory of Mind In Communicating Agents [paper] [website] [arxiv]
  5. 2024: Bortoletto, M., Ruhdorfer, C., Shi, L. & Bulling, A. (2024). Explicit Modelling of Theory of Mind for Belief Prediction in Nonverbal Social Interactions. Proc. European Conference on Artificial Intelligence (ECAI), pp. 1–7, 2024. (oral) [paper] [website] [arxiv]
  6. 2024: Bortoletto, M., Ruhdorfer, C., Shi, L. & Bulling, A. (2024). Benchmarking Mental State Representations in Language Models. ICML Workshop on Mechanistic Interpretability. [paper] [website] [arxiv]
  7. 2024: Bortoletto, M., Ruhdorfer, C., Abdessaied, A., Shi, L. & Bulling, A. (2024). Limits of Theory of Mind Modelling in Dialogue-Based Collaborative Plan Acquisition. Proc. 62nd Annual Meeting of the Association for Computational Linguistics (ACL), pp. 1–16, 2024. [paper] [website] [arxiv] [acl]
  8. 2024: Wang, Y., Jiang, Y., Hu, Z., Ruhdorfer, C., Bhâce, M., & Bulling, A. (2024). VisRecall++: Analysing and Predicting Visualisation Recallability from Gaze Bahaviour. Proc. ACM on Human-Computer Interaction (PACM HCI), 1–11. [paper] [website] [acm]
  9. 2020: Ruhdorfer, C., & Schulz, S. (2020). Efficient Implementation of Large-Scale Watchlists. PAAR+SC2 @ International Joint Conference on Automated Reasoning (IJCAR) (pp. 120-133). [paper] [website]
I sometimes collaborate with others to work on topics not related to my core research interests:
  1. 2025: Hofmeyer, P. E., Burghaus, H., Ruhdorfer, C., Oswald, J., Bulling, A., Herdrich, G. (2025) Integration of Machine Learning in High-Enthalpy Plasma Spectroscopy. Proc. International Conference on Flight vehicles, Aerothermodynamics and Re-entry (FAR), pp. 1–8 [paper] [website]

Preprints

  1. 2025: Ruhdorfer, C., Bortoletto, M., Oei, V., Penzkofer, A. & Bulling, A. (2025). Unsupervised Partner Design Enables Robust Ad-Hoc Teamwork. [arxiv]

Conferences, Workshops and Other Academic Meetings

Awards

Teaching

For details please refer to here.

Supervision

I regulary supervise Master and Bachelor theses, see here for a full list and if interested our open projects page. To this day I was or am involved in supervising a total of 3 students in their Master Thesis and a total of 8 students in various group projects offered through some of our lectures.

News

Two papers accepted at EMNLP 2025

I am very happy to announce that we have two papers accepted at EMNLP 2025! Both works deal with Theory of Mind and Large Language Models. In the first we study how LLMs represent beliefs (findings) and in the second we propose a new benchmark for evaluating the Theory of Mind abilities of LLMs (main conference).

Matteo Bortoletto, Constantin Ruhdorfer, Lei Shi, Andreas Bulling Brittle Minds, Fixable Activations: Understanding Belief Representations in Language Models Findings of Empirical Methods in Natural Language Processing, 2025.

Matteo Bortoletto, Constantin Ruhdorfer, Andreas Bulling ToM-SSI: Evaluating Theory of Mind in Situated Social Interactions Proc. Empirical Methods in Natural Language Processing, 2025.

Paper Accepted at TMLR

Our new work on evaluating agents’ ability to cooperate with new partners in unknown environments has been accepted at TMLR. The paper is titled The Overcooked Generalisation Challenge: Evaluating Cooperation with Novel Partners in Unknown Environments Using Unsupervised Environment Design.

Paper Accepted at ToM IJCAI 2025

I am very happy to announce that we have a paper accepted at ToM IJCAI 2025. Our work will be featured as a long oral talk at the workshop. The paper is titled The Yōkai Learning Environment: Tracking Beliefs Over Space and Time and proposes a new multi-agent RL environment for zero-shot coordination and Theory of Mind research.

Accepted into the 2025 Cooperative AI Foundation Summmer School

Super excited to announce that I have been accepted into the 2025 edition of the Cooperative AI foundation’s summer school in Marlow, UK.

Attended the CoGenAI Summer School

I had a blast attending the ELLIS CoGenAI Summer School last week where I talked about our recent preprint “The Overcooked Generalisation Challenge” available here. For more details you can also see my LinkedIn post