CV

PhD in Neuroscience and Machine Learning.

Wellcome Trust PhD programme, Faculty of Engineering Mathematics, University of Bristol, UK.

• Applications of deep reinforcement learning to understanding human motor learning, especially with policy gradient methods (60% of PhD).
• Purely machine learning projects on improving value-based estimation methods in reinforcement learning (40% of PhD).

Contract education in Data Science and Knowledge Engineering.

Maastricht University, Netherlands.

• Undertook fundamental courses in mathematics and computer science at bachelor and master levels (average grade achieved 8.4/10).

Master of Research (MRes) in Cognitive Neuroscience.

University College London (UCL), UK.

• Awarded with Distinction; Dean's list award for top 5% of the Brain faculty.

Bachelor of Science (BSc) in Psychology.

Bath Spa University, UK.

• Awarded First-class; British Psychology Association award for top graduating student.

Senior Postdoctoral Associate in Protein Language Models & Generative Protein Design.

Centre for Genomic Regulation (Ferruz group), Spain.

• Led design and execution of large-scale pre-training of protein language models (ProtGPT3 project, up to 10B parameters), across multi-node, multi-GPU systems (40+ GPUs) using PyTorch Distributed, DeepSpeed, and Hugging Face Accelerate.
• Led post-training (SFT, RLHF) of large protein and genomic (e.g., evo2) language models for target bio properties.
• Developed large-scale evaluation pipelines for protein language model generations, integrating structure prediction (AlphaFold, ESMFold, Boltz), evolutionary consistency checks (HMM profiles, MSA-based metrics), and sequence-level scoring (ESM, ProteinMPNN) to assess structural validity and design quality at scale.

Visiting Postdoctoral Research Associate in Neuroscience and Machine Learning.

University of Oxford, UK, under Rui Ponte Costa.

• April - December 2024.

Visiting PhD Student in Neuroscience and Machine Learning.

University of Oxford, UK.

Teaching Assistant.

University of Bristol, UK.

• Engineering Mathematics 1, Computational Neuroscience.

Paid Student Assistantship in Deep Reinforcement Learning and Human Vision.

Maastricht University, Netherlands.

• Feb - April 2019. Working on a well-known deep RL vision architecture to subsequently model human gaze (repo).

Assistant Data Analyst.

Annex Clinical, New York, USA.

• Feb - April 2016. Analysis of a large clinical dataset to investigate the relation between anhedonia and depression in relation to clinical trial drop-out (see article).

• Garibbo, M., Boxo, G., Middendorf, L., Stocco, F., & Ferruz, N. ProtGPT3: an Open-source family of Promptable and Aligned Protein Language Models. NeurIPS.

• Stocco*, F., Garibbo*, M., & Ferruz, N. Steering generative models for protein design: Aligning and conditioning strategies. Current Opinion in Structural Biology, 98, 103250.
• Moberg, S., Garibbo*, M., Mazo*, C., Gilad, A., Schmitz, D., Costa, R. P., ... & Takahashi, N. Distinct roles of cortical layer 5 subtypes in associative learning. Nature Communications, 17(1), 2648.

• Garibbo, M., Robeyns, M., & Aitchison, L. Taylor TD-learning. Advances in Neural Information Processing Systems (NeurIPS), 37.

• Garibbo, M., Boxo, G., Middendorf, L., Stocco, F., & Ferruz, N. ProtGPT3: an Open-source family of Promptable and Aligned Protein Language Models. Eric and Wendy Schmidt Symposium on Biomedical Science and AI, Broad Institute of MIT & Harvard, USA.

• Garibbo, M., Ludwig, C., Lepora, N., & Aitchison, L. What can deep reinforcement learning tell us about human motor learning and vice-versa? Neuromatch Conference, December 2021, Online.

• Garibbo, M. and Wierdak, E. The effect of local environmental context changes on intentional and unintentional memory retrieval. British Psychology Association South West Undergraduate Conference, March 2016, University of the West of England, Bristol.

• Garibbo, M., Robeyns, M., & Aitchison, L. Taylor TD-learning. NeurIPS, USA.

• Zanzi, M., Garibbo, M., Tavano, A., & Saponati, M. RNN reconstruction of mouse latent neural dynamics. Neuromatch Conference, Online.
• Garibbo, M., Ludwig, C., Lepora, N., & Aitchison, L. What deep reinforcement learning tells us about human motor learning and vice-versa. CSHL From Neuroscience to Artificially Intelligent conference, New York.

• MINECO Colaboración Público-Privada for a Deep Learning Scientist in protein design.

• Wellcome Trust 8-month Transition Fund for a Postdoctoral position.

• Wellcome Trust 4-year PhD Scholarship.

• MN5 EuroHPC Development Access Call: 14,000 GPU hours.

• MN5 Red Española de Supercomputación fast-track low priority GPU access.

• Dean's list award for top 5% graduating student of the UCL Brain faculty.

• British Psychology Association award for the top psychology graduating student of Bath Spa University.