Raphael Avalos
Raphael Avalos

PhD Candidate & Research Intern

About Me

Raphael Avalos is a PhD student at the AI Lab of Vrije Universiteit Brussel (VUB), supervised by Professors Ann Nowé and Diederik Roijers. Raphael’s PhD research addresses the challenges of partial observability in both single-agent and multi-agent reinforcement learning (RL), investigating how to leverage additional data during the learning phase and exploring the potential of communication at the execution stage to enhance decision-making under uncertainty. His main interests lie in model-based RL and exploration strategies.

Download CV
Interests
  • Artificial Intelligence
  • Reinforcement Learning
  • POMDPs
  • Multi-Agent RL
  • Model Based RL
  • Exploration
  • Large Languages Models
Recent News

Experience

  1. Research Intern

    Cohere AI
  2. Program Chair

    Adaptive and Learning Agents Workshop (ALA) at AAMAS
  3. Local Chair

    European Workshop on Reinforcement Learning (EWRL)
  4. Visiting Researcher

    Delft University of Technology
  5. Research Intern

    Vrije Universiteit Brussel (VUB) - AI Lab
  6. Research Intern

    INRIA - SequeL Team (now Scool)

Education

  1. PhD in Artificial Intelligence (ongoing)

    Vrije Universiteit Brussel
  2. MSc in Applied Mathematics for AI (MVA)

    Ecole Normale Superieur de Cachan
  3. MSE in Computer Science

    Telecom Paris
Awards
PhD Fellowship Fundamental Research
FWO - Research Foundation Flanders ∙ November 2020
Best Paper Award
Benelux Conference on Artificial Intelligence (BNAIC) ∙ September 2023
Recent Publications
(2024). Dynamic Size Message Scheduling for Multi-Agent Communication under Limited Bandwidth. IEEE Transactions on Mobile Computing.
(2024). Online Planning in POMDPs with State-Requests. RLJ - Reinforcement Learning Journal.
(2024). The Wasserstein Believer: Learning Belief Updates for Partially Observable Environments through Reliable Latent Space Models. The Twelfth International Conference on Learning Representations.
(2023). Laser Learning Environment: A new environment for coordination-critical multi-agent tasks. The 35th Benelux Conference on Artificial Intelligence and the 32th Belgian Dutch Conference on Machine Learning.
(2023). Local Advantage Networks for Multi-Agent Reinforcement Learning in Dec-POMDPs. Transactions on Machine Learning Research.