Papers

On the computational principles underlying human exploration
Lior Fox, Ohad Dan, Yonatan Loewenstein
PsyArXiv preprint 2023

Reinforcement Learning with Large Action Spaces for Neural Machine Translation
Asaf Yehudai, Leshem Choshen, Lior Fox, Omri Abend
International Conference on Computational Linguistics (COLING) 2022

Exploration: from machines to humans
Lior Fox*, Ohad Dan*, Lotem Elber-Dorozko*, Yonatan Loewenstein
Current Opinion in Behavioral Sciences 2021

On the Weaknesses of Reinforcement Learning for Neural Machine Translation
Leshem Choshen, Lior Fox, Zohar Aizenbud, Omri Abend
International Conference on Learning Representations (ICLR) 2020

DORA The Explorer: Directed Outreaching Reinforcement Action-Selection
Lior Fox*, Leshem Choshen*, Yonatan Loewenstein
International Conference on Learning Representations (ICLR) 2018

Talks and presentations

Long-term consequences of actions affect human exploration in structured environments
Lior Fox, Ohad Dan, Gal Yarden, Yonatan Loewenstein
Computational and Systems Neuroscience (Cosyne) 2022

Maximum Entropy Approach for Optimal Exploration
Lior Fox, Yonatan Loewenstein
Reinforcement Learning and Decision Making (RLDM) 2019

In search of optimal exploration
Lior Fox, Yonatan Loewenstein
Computational and Systems Neuroscience (Cosyne) 2019

Propagating Directed Exploration in Model-Free Reinforcement Learning
Lior Fox*, Leshem Choshen*, Yonatan Loewenstein
Reinforcement Learning and Decision Making (RLDM) 2017

Book (upcoming)

The Dynamics of Computation in Neuronal Networks
Yonatan Loewenstein, Lior Fox, Itamar Landau, and Gianluigi Mongillo
Expected publication late 2023, MIT Press (accepted)

We are at the final steps of writing a computational neuroscience book. The book is intended to serve teachers and students of graduate-level courses in modelling networks dynamics.

Thesis

My PhD thesis, Exploration in Complex Environments: Computational Modeling and Human Behavior, including an otherwise unpublished chapter on Maximum Entropy exploration.

* denotes equal contribution

Jekyll template design adapted from researcher theme