PeopleofSuperintelligence

Researcher

Sergio Guadarrama

Sergio Guadarrama is an AI researcher whose work spans computer vision, deep-learning systems, and reinforcement learning. He earned bachelor's and PhD degrees at the Technical University of Madrid and received its Best Doctoral Dissertation Award for 2006–2007. At Berkeley, he helped develop Caffe and received the 2017 Everingham Prize for it; at Google, he contributed to TensorFlow training libraries and self-supervised video tracking. His later papers address predictive information and multi-game decision transformers, and he co-authored the A2Perf autonomous-agent benchmark. In May 2026, he is among AlphaEvolve's users.

Works
94
Citations
29,091
h-index
31

Publication numbers from OpenAlex, as of 15 October 2026.

Timeline

  1. 2006–2007

    A doctoral dissertation award

    He receives the Best Doctoral Dissertation Award from the Technical University of Madrid for 2006–2007.

  2. 2014

    Caffe at Berkeley

    Before Google, he works as a researcher at UC Berkeley and helps develop Caffe. The 2014 paper presents Caffe as a framework for deep-learning algorithms.

  3. 2016–2017

    TensorFlow and Caffe

    In 2016, he contributes to TensorFlow's training libraries. In 2017, he receives the Everingham Prize for Caffe, an open-source deep-learning framework.

  4. 2018

    Self-supervised video tracking

    By 2018, he holds bachelor's and PhD degrees from the Technical University of Madrid. He is part of Google's core team for Self-Supervised Tracking via Video Colorization.

  5. 2020

    Predictive information in RL

    In 2020, he co-authors a paper on predictive information for reinforcement learning.

  6. 2022

    Multi-Game Decision Transformers

    In 2022, he co-authors Multi-Game Decision Transformers, a multi-game transformer study.

  7. 2025

    The A2Perf benchmark

    In 2025, he co-authors A2Perf, a benchmark for real-world autonomous agents.

  8. May 2026

    Using AlphaEvolve

    In May 2026, he is among AlphaEvolve's users.

Sources and credits

Photo: We have no published photo of Sergio Guadarrama yet. Send us a free-licensed photo.

  1. XVIII Conference of the Spanish Association for Artificial Intelligence sci2s.ugr.es. Checked 15 October 2026.
  2. Reinforcement Learning dasci.es. Checked 15 October 2026.
  3. Caffe: Convolutional Architecture for Fast Feature Embedding arxiv.org. Checked 15 October 2026.
  4. Announcing TensorFlow 0.8 – now with distributed computing support! research.google. Checked 15 October 2026.
  5. Self-Supervised Tracking via Video Colorization research.google. Checked 15 October 2026.
  6. Predictive Information Accelerates Learning in RL proceedings.neurips.cc. Checked 15 October 2026.
  7. Multi-Game Decision Transformers proceedings.neurips.cc. Checked 15 October 2026.
  8. A2Perf: Real-World Autonomous Agents Benchmark arxiv.org. Checked 15 October 2026.
  9. AlphaEvolve: How our Gemini-powered coding agent is scaling impact across fields deepmind.google. Checked 15 October 2026.