Me at the church of San Pedro de Atacama.

I am a Postdoctoral Research Fellow at the Sensorimotor Learning and Decision-Making Group working with Daniel A. Braun and a researcher at ORAND.

I'm a member of the new Max-Planck Institute for Intelligent Systems and the Max-Planck Institute for Biological Cybernetics in Tübingen, Germany.

I did my PhD in Autonomous Agents under the supervision of Zoubin Ghahramani at the University of Cambridge and my BSc at the University of Chile.

Two Fully-Funded PhD Positions in Computational Neuroscience (Positions filled)

I work on the foundations of autonomous agents, and I study the links between:

• information,
• control,
• learning,
• causality,
• complexity,
• physics,
• and economics.

In particular, I am interested in:

• resource-bounded adaptive control and bounded rationality,
• and the foundations of abstraction and structural learning.

If you want to get a clear idea of what I do, then you can:

• read my research statement,
• read my thesis,
• or have a look at my publications.

I am not interested in particular algorithms: rather, I'm interested in the unifying ideas.

• Bayesian Causal Induction
• Bayesian Control Rule

• 17th December 2011: The talk “Bayesian Causal Induction” can be found here.
• 15th November 2011: I'm giving the talk “Bayesian Causal Induction” at the 2011 NIPS Workshop on Philosophy and Machine Learning, Sierra Nevada.
• 2nd November 2011: Paper “Bayesian Causal Induction”, ArXiv 1111.0708.
• 26th July 2011: Started working as a postdoctoral fellow at the Max Planck Institute for Biological Cybernetics.
• 21th June 2011: “A Unified Framework for Resource-Bounded Agents Interacting with an Unknown Environment”, final draft, PhD dissertation.
• 4th May 2011: Papers “Information, Utility and Bounded Rationality” and “Reinforcement Learning and the Bayesian Control Rule” accepted at the fourth conference on artificial general intelligence.
• 11th April 2011: Talk “A Minimum Relative Entropy Principle for Learning and Acting”, Symposium of the Max Planck Institute for Intelligent Systems, Tübingen, Germany.
• more...

$$ P(a_{t+1}|\hat{a}_{1,\ldots,t},o_{1,\ldots,t}) = \int P(a_{t+1}|\theta, \hat{a}_{1,\ldots,t},o_{1,\ldots,t}) P(\theta|\hat{a}_{1,\ldots,t},o_{1,\ldots,t}) \, d\theta $$