Alejandro Escontrela
Alejandro Escontrela studies reinforcement learning, unsupervised learning and robotics. His work includes learning from behavioural priors and video and transferring skills to robots. Escontrela co-authored the RSS 2022 paper Autonomously Untangling Long Cables. The team combined perception with robot actions to manipulate cable tangles. He released Video Prediction Rewards in May 2023, and it was accepted at NeurIPS that year. The project connects predictive video models with reinforcement-learning rewards. Escontrela joined the Google DeepMind Robotics team as a research scientist on 9 February 2026 with Jie Tan. His research interests include behavioural priors and efficient reinforcement learning.
- Now at
- Google DeepMind
- Works
- 13
- Citations
- 225
- h-index
- 7
Publication numbers from OpenAlex, as of 14 October 2026.
Timeline
-
2022
Cable manipulation
Escontrela co-authored the RSS 2022 paper Autonomously Untangling Long Cables. The team combined perception with robot actions to manipulate cable tangles.
-
2023
Video prediction rewards
He released Video Prediction Rewards in May 2023, and it was accepted at NeurIPS that year. The project connects predictive video models with reinforcement-learning rewards.
-
2026
DeepMind robotics
Escontrela joined the Google DeepMind Robotics team as a research scientist on 9 February 2026 with Jie Tan. His research interests include behavioural priors and efficient reinforcement learning.
Sources and credits
Photo: www.escontrela.me; creator not stated. License: Not stated by source. Source: escontrela.me. Cropped to a square and resized. Ask us to remove or replace this photo.