Marco Baroni

Research Scientist

I am a research scientist on the Facebook Artificial Intelligence Research (FAIR) team, joining in November 2016. I received my PhD in theoretical linguistics from the University of California, Los Angeles in 2000. After various experiences in research and industry, I joined the faculty of the University of Trento (Italy) in 2006, as associate professor at the Center for Mind/Brain Sciences. My work there focused on general methods for natural language understanding that rely on raw input data, such as large amounts of text and images. My work was partially supported by a Google Research Award and an ERC Starting grant.

My focus at FAIR is on methods to train machines to interact with humans (and with each other) through natural language. To address this challenge in a general way, we need to tackle issues such as the following: where to find (or how to generate) data to train conversational machines in the right interactive setup; how to make them adapt fast to the constantly changing state of conversations; how to make them aware of the real-world context in which a conversation is taking place; how to make them understand what is part of the common-ground knowledge shared by the participants in the conversation; and so on. I am interested in all of these aspects and more.


Methods to train machines to interact with humans through natural language

Latest Publications

ICML - August 13, 2020

Entropy Minimization In Emergent Languages

Eugene Kharitonov, Rahma Chaabouni, Diane Bouchacourt, Marco Baroni

ACL - July 4, 2020

Compositionality and Generalization in Emergent Languages

Rahma Chaabouni, Eugene Kharitonov, Diane Bouchacourt, Emmanuel Dupoux, Marco Baroni

ICLR - April 25, 2020

Permutation Equivariant Models for Compositional Generalization in Language

Jonathan Gordon, David Lopez-Paz, Marco Baroni, Diane Bouchacourt

NeurIPS - December 8, 2019

Anti-efficient encoding in emergent communication

Rahma Chaabouni, Eugene Kharitonov, Emmanuel Dupoux, Marco Baroni

EMNLP - November 2, 2019

EGG: a toolkit for research on Emergence of lanGuage in Games

Eugene Kharitonov, Rahma Chaabouni, Diane Bouchacourt, Marco Baroni

ACL - July 29, 2019

Word-order biases in deep-agent emergent communication

Rahma Chaabouni, Eugene Kharitonov, Alessandro Lazaric, Emmanuel Dupoux, Marco Baroni

ACL - July 26, 2019

On the Distribution of Deep Clausal Embeddings: A Large Cross-linguistic Study

Damián E. Blasi, Ryan Cotterell, Lawrence Wolf-Sonkin, Sabine Stoll, Balthasar Bickel, Marco Baroni

Philosophical Transactions of the Royal Society B - July 3, 2019

Linguistic generalization and compositionality in modern artificial neural networks

Marco Baroni

NAACL - June 2, 2019

The emergence of number and syntax units in LSTM language models

Yair Lakretz, Germán Kruszewski, Theo Desbordes, Dieuwke Hupkes, Stanislas Dehaene, Marco Baroni

IJCAI-ECAI Workshop - January 30, 2019

Memorize or generalize? Searching for a compositional RNN in a haystack

Adam Liška, Germán Kruszewski, Marco Baroni

EMNLP 2018 - November 2, 2018

Jump to better conclusions: SCAN both left and right

Joost Bastings, Marco Baroni, Jason Weston, Kyunghyun Cho, Douwe Kiela

EMNLP 2018 - November 1, 2018

How agents see things: On visual representations in an emergent language game

Diane Bouchacourt, Marco Baroni

ACL 2018 - July 18, 2018

What you can cram into a single $&!#* vector: Probing sentence embeddings for linguistic properties

Alexis Conneau, Germán Kruszewski, Guillaume Lample, LoÏc Barrault, Marco Baroni

NAACL 2018 - June 1, 2018

Colorless Green Recurrent Networks Dream Hierarchically

Kristina Gulordava, Piotr Bojanowski, Edouard Grave, Tal Linzen, Marco Baroni

IWCS 2017 - October 8, 2017

Living a Discrete Life in a Continuous World: Reference in Cross-Modal Entity Tracking

Gemma Boleda, Sebastian Pado', Nghia The Pham, Marco Baroni

ICLR - April 24, 2017

Multi-Agent Cooperation and the Emergence of (Natural) Language

Angeliki Lazaridou, Alexander Peysakhovich, Marco Baroni

ICLR - April 24, 2017

CommAI: Evaluating the First Steps Towards a Useful General AI

Marco Baroni, Armand Joulin, Allan Jabri, Germán Kruszewski, Angeliki Lazaridou, Klemen Simonic, Tomas Mikolov

ArXiv - November 25, 2015

A Roadmap Towards Machine Intelligence

Tomas Mikolov, Armand Joulin, Marco Baroni