Skip to main content

Multimodal Agents Grounded via Interactive Communication

Searching for OpenAIRE data...

Publications

On the realization of compositionality in neural networks

Author(s): J. Baan, j. Leible, M. Nikolaus, D. Rau, T. Baumgärtner, D. Hupkes and E. Bruni
Published in: Proceedings of the Second BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP, 2019

Transcoding compositionally: using attention to find more generalizable solutions

Author(s): Kris Korrel, Dieuwke Hupkes, Verna Dankers, Elia Bruni
Published in: 2019

Location Attention for Extrapolation to Longer Sequences

Author(s): Yann Dubois, Gautier Dagan, Dieuwke Hupkes, Elia Bruni
Published in: 2020

Assessing incrementality in sequence-to-sequence models

Author(s): D Ulmer, D Hupkes, E Bruni
Published in: Proceedings of the 4th Workshop on Representation Learning for NLP (RepL4NLP-2019), 2019

Mastering emergent language: learning to guide in simulated navigation

Author(s): Mathijs Mul, Diane Bouchacourt, Elia Bruni
Published in: 2019

Incentivizing the Emergence of Grounded Discrete Communication Between General Agents

Author(s): Thomas A. Unger, Elia Bruni
Published in: 2020

Co-evolution of language and agents in referential games

Author(s): Gautier Dagan, Dieuwke Hupkes, Elia Bruni
Published in: 2020

Exploiting Language Instructions for Interpretable and Compositional Reinforcement Learning

Author(s): Michiel van der Meer, Matteo Pirotta, Elia Bruni
Published in: 2020

Compositional properties of emergent languages in deep learning

Author(s): Bence Keresztury, Elia Bruni
Published in: 2020

Internal and external pressures on language emergence: least effort, object constancy and frequency

Author(s): Diana Rodríguez Luna, Edoardo Maria Ponti, Dieuwke Hupkes, Elia Bruni
Published in: 2020

Compositionality Decomposed: How do Neural Networks Generalise?

Author(s): Dieuwke Hupkes, Verna Dankers, Mathijs Mul, Elia Bruni
Published in: Journal of Artificial Intelligence Research, Issue 67, 2020, Page(s) 757-795, ISSN 1076-9757
DOI: 10.1613/jair.1.11674