I am currently a Postdoctoral Researcher at the Max Planck Institute for Security and Privacy (MPI-SP), Germany with Director Meeyoung Cha. Previously, I was a Postdoctoral Researcher at the Institute for Basic Science, South Korea. I received my PhD from KAIST and my Bachelor’s and Master’s degrees from the Federal University of Rio de Janeiro, Brazil.

Research Interests

My research vision is to develop safe artificial intelligence algorithms for multi-agent systems that advance our knowledge on how humans and machines learn, interact, and collaborate.

My general research interests include Reinforcement Learning, Multi-Agent Systems, and Robotics. From these, I am particularly interested in:

  • Reinforcement Learning: continual learning, multi-agent RL, improving sampling efficiency and performance by developing novel credit assignment and exploration methods, and goal-conditioned RL.
  • Multi-Agent Systems: development of multi-agent RL algorithms in collaborative and collaborative-competitive environments, modeling of continual multi-agent frameworks.
  • Robotics: end-to-end control, safety and risks of robotic systems, and hardware security.

Selected Publications

  1. Self-Evolving Agents Should Build Internal and External Models of the World
    Vecchietti, L. F., Wijaya, B. N., Dong, W., Cha, M.
    Advances in Neural Information Processing Systems (NeurIPS), 2026
  2. Mitigating Plasticity Loss through Architectural Design in Continual Learning
    Koeppe, N., Vecchietti, L. F. et al.
    Proceedings of the International Conference on Machine Learning (ICML), 2026
  3. Textual Supervision Enhances Geospatial Representations in Vision-Language Models
    Locatelli, M. S., Tonucci, F., Kwon, J., Vecchietti, L. F. et al.
    Proceedings of the International Conference on Machine Learning (ICML), 2026
  4. Machine Behavior in Relational Moral Dilemmas: Moral Rightness, Predicted Human Behavior, and Model Decisions
    Kim, K., Kwon, J., Vecchietti, L. F. et al.
    ACL Findings, 2026 — Co-corresponding author
    paper code
  5. Interpretable Machine Learning for Protein Science: Structure, Function, and Interactions
    Vecchietti, L. F. et al.
    ACM Computing Surveys, 2026
    paper
  6. Dropouts in Confidence: Moral Uncertainty in Human-LLM Alignment
    Kwon, J., Vecchietti, L. F. et al.
    Proceedings of the AAAI Conference on Artificial Intelligence, AI Alignment Track, 2026
    paper
  7. Artificial intelligence-driven computational methods for antibody design and optimization
    Vecchietti, L. F., Wijaya, B. N. et al.
    mAbs, 2025
    paper
  8. Robust Optimization in Protein Fitness Landscapes Using Reinforcement Learning in Latent Space
    Vecchietti*, L. F., Lee*, M. et al.
    Proceedings of the International Conference on Machine Learning (ICML), 2024 — Spotlight Poster, co-first authors
    paper code
  9. AI World Cup: Robot Soccer-Based Competitions
    Hong, C., Jeong, I., Vecchietti, L. F. et al.
    IEEE Transactions on Games, 2021 — Main simulation environment for the AI World Cup competitions
    paper
  10. Sampling Rate Decay in Hindsight Experience Replay for Robot Control
    Vecchietti, L. F., Seo, M. et al.
    IEEE Transactions on Cybernetics, 2020
    paper
  11. Rewards Prediction Based Credit Assignment for Reinforcement Learning with Sparse Binary Rewards
    Seo, M., Vecchietti, L. F. et al.
    IEEE Access, 2019 — Selected as featured research in KAIST Breakthroughs Magazine 2020
    paper

Diversity Statement

I am committed to increasing diversity in AI research. If you are from an underrepresented group and need research mentoring and guidance, feel free to contact me for a mentoring session.

Personal Interests

My personal interests include traveling, sports (football, kickboxing, skate, snowboard, surf, surfskate), music (classical guitar, cavaquinho) and reading (fantasy, science fiction, mountaineering).