Alvaro Velasquez

Papers

Total Citations

H-Index

About

Álvaro Velásquez is a researcher at the forefront of reinforcement learning (RL), with a focus on making sequential decision-making more efficient, scalable, and practically deployable. His work addresses one of the central challenges in modern AI: bridging the gap between theoretical RL capabilities and real-world application. Velásquez has contributed meaningfully to curriculum learning for RL agents, demonstrating how automaton-guided frameworks can automatically generate reward functions from logical task specifications to accelerate training in complex environments. His research on sim-to-real transfer — surveying how RL policies trained in simulation can be reliably deployed in the physical world — situates him at a critical intersection of robotics, policy generalization, and foundation models. He has also explored the integration of Large Language Models into RL pipelines, developing dynamic task sampling methods that leverage LLM-generated sub-goals to guide agent learning without restrictive assumptions. Though his citation counts are still growing — reflecting the recency of his contributions — his work spans timely and high-impact themes that are shaping the next generation of intelligent, autonomous systems. Velásquez represents an emerging voice in the RL community with a broad and forward-looking research vision.