Remo Sasso

Applied Scientist II at Amazon · RL researcher

Remo Sasso presenting research at a conference

$ whoami

I work on reinforcement learning, exploration, model-based RL, posterior sampling, and foundation models as world models or decision-making priors. I completed my PhD in Artificial Intelligence at Queen Mary University of London, supervised by Paulo Rauber.

Selected Publications

Full list on Google Scholar.

2026
Foundation Models as World Models: A Foundational Study in Text-Based GridWorlds

IEEE Conference on Games (CoG) · Sasso, Conserva, Jeurissen, Rauber

2025
Exploration with Foundation Models: Capabilities, Limitations, and Hybrid Approaches

NeurIPS 2025 FoRLM Workshop · Sasso, Conserva, Jeurissen, Rauber · Paper

2023
Posterior Sampling for Deep Reinforcement Learning

ICML · Sasso, Conserva, Rauber · Paper · Code

2023
Multi-Source Transfer Learning for Deep Model-Based Reinforcement Learning

TMLR · Sasso, Sabatelli, Wiering · Paper · Code

Recent Logs

2025-now Applied Scientist II, Amazon. Broad applied ML systems, evaluation, experimentation, and production-quality implementation.
2025 Applied Scientist Intern, Amazon Science. Recommendation, personalization, reinforcement learning, and empirical ML evaluation.
2023-25 Lead AI Developer, xDNA / Aletheia. Agentic retrieval, fact-checking, multilingual analysis, and evidence synthesis.
2021-25 PhD Researcher, Queen Mary University of London. RL, exploration, model-based RL, posterior sampling, and foundation-model priors.