Autonomous Linguistic Emergence in neural Networks

Informations projet

ALiEN

N° de convention de subvention: 101019291

DOI

10.3030/101019291

Date de signature de la CE 26 Mai 2021

Date de début 1 Janvier 2022

Date de fin 31 Decembre 2026

Financé au titre de

• EXCELLENT SCIENCE - European Research Council (ERC)

Coût total

€ 2 489 541,00

Contribution de l’UE

€ 2 489 541,00

Coordonné par

UNIVERSIDAD POMPEU FABRA
Spain

CORDIS fournit des liens vers les livrables publics et les publications des projets HORIZON.

Les liens vers les livrables et les publications des projets du 7e PC, ainsi que les liens vers certains types de résultats spécifiques tels que les jeux de données et les logiciels, sont récupérés dynamiquement sur OpenAIRE .

Publications

Conference proceedings (19)

The emergence of numerical representations in communicating artificial agents

Auteurs: Daniela Mihai, Lucas Weber, Francesca Franzon
Publié dans: Proceedings of Evolang XVI, 2026, Page(s) 308-315
Éditeur: The Evolution of Language Conferences

Cross-domain image captioning with discriminative finetuning.

Auteurs: R. Dessì, M. Bevilacqua, E. Gualdoni, N. Rakotonirina, F. Franzon, M. Baroni
Publié dans: Proceedings of CVPR 2023 (IEEE/CVF Conference on Computer Vision and Pattern Recognition), 2023, Page(s) 6935-6944
Éditeur: Computer Vision Foundation

Towards hate speech detection in low-resource languages: Comparing ASR to acoustic word embeddings on Wolof and Swahili

Auteurs: Jacobs, C., Rakotonirina, N.C., Chimoto, E.A., Bassett, B.A., Kamper, H.
Publié dans: Proc. INTERSPEECH 2023, 2023
Éditeur: ISCA
DOI: 10.21437/interspeech.2023-421

On the Correspondence between Compositionality and Imitation in Emergent Neural Communication

Auteurs: E. Cheng, M. Rita, T. Poibeau
Publié dans: Findings of the Association for Computational Linguistics: ACL 2023, 2023, Page(s) 12432–12447
Éditeur: ACL

Referential communication in heterogeneous communities of pre-trained visual deep networks.

Auteurs: M. Mahaut, F. Franzon, R. Dessì, M. Baroni
Publié dans: Proc. of the 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2023), 2023, Page(s) 2619-2621
Éditeur: International Foundation for Autonomous Agents and Multiagent Systems

Not a nuisance but a useful heuristic: Outlier dimensions favor frequent tokens in language models

Auteurs: Iuri Macocco, Nora Graichen, Gemma Boleda, Marco Baroni
Publié dans: Proceedings of the 8th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP, 2025, Page(s) 109-136
Éditeur: Association for Computational Linguistics
DOI: 10.18653/v1/2025.blackboxnlp-1.6

Geometric Signatures of Compositionality Across a Language Model’s Lifetime

Auteurs: Jin Hwa Lee, Thomas Jiralerspong, Lei Yu, Yoshua Bengio, Emily Cheng
Publié dans: Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2025, Page(s) 5292-5320
Éditeur: Association for Computational Linguistics
DOI: 10.18653/v1/2025.acl-long.265

Communication breakdown: On the low mutual intelligibility between human and neural captioning

Auteurs: R. Dessì, E. Gualdoni, F. Franzon, G. Boleda, M. Baroni
Publié dans: Proceedings of EMNLP 2022 (Conference on Empirical Methods in Natural Language Processing), 2022, Page(s) 7998-8007
Éditeur: ACL

Factual Confidence of LLMs: on Reliability and Robustness of Current Estimators

Auteurs: Matéo Mahaut, Laura Aina, Paula Czarnowska, Momchil Hardalov, Thomas Müller, Lluis Marquez
Publié dans: Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2024, Page(s) 4554-4570
Éditeur: Association for Computational Linguistics
DOI: 10.18653/v1/2024.acl-long.250

Evil twins are not that evil: Qualitative insights into machine-generated prompts

Auteurs: Nathanaël Carraz Rakotonirina, Corentin Kervadec, Francesca Franzon, Marco Baroni
Publié dans: Proceedings of the 8th BlackboxNLP Workshop: Analyzing and Interpreting Neural Networks for NLP, 2025, Page(s) 48-68
Éditeur: Association for Computational Linguistics
DOI: 10.18653/v1/2025.blackboxnlp-1.3

Bridging information-theoretic and geometric compression in language models

Auteurs: Cheng, Emily; Kervadec, Corentin; Baroni, Marco
Publié dans: Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, 2023
Éditeur: Association for Computational Linguistics

MemoryPrompt: A light wrapper to improve context tracking in pre-trained language models

Auteurs: Rakotonirina, Nathanael; Baroni, Marco
Publié dans: Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), 2024
Éditeur: Association for Computational Linguistics

Evidence from fMRI Supports a Two-Phase Abstraction Process in Language Models

Auteurs: Richard Antonello, Emily Cheng
Publié dans: Proceedings of UniReps@NeurIPS 2024, 2024
Éditeur: OpenReview/Online

Prediction Hubs are Context-Informed Frequent Tokens in LLMs

Auteurs: Beatrix Miranda Ginn Nielsen, Iuri Macocco, Marco Baroni
Publié dans: Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2025, Page(s) 23715-23745
Éditeur: Association for Computational Linguistics
DOI: 10.18653/v1/2025.acl-long.1156

Unnatural language processing: How do language models handle machine-generated prompts?

Auteurs: Kervadec, Corentin; Franzon, Francesca; Baroni, Marco
Publié dans: Findings of EMNLP, 2023
Éditeur: Association for Computational Linguistics
DOI: 10.18653/v1/2023.findings-emnlp.959

From Tools to Teammates: Evaluating LLMs in Multi-Session Coding Interactions

Auteurs: Nathanaël Carraz Rakotonirina, Mohammed Hamdy, Jon Ander Campos, Lucas Weber, Alberto Testoni, Marzieh Fadaee, Sandro Pezzelle, Marco Del Tredici
Publié dans: Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2025, Page(s) 19609-19642
Éditeur: Association for Computational Linguistics
DOI: 10.18653/v1/2025.acl-long.964

Correct, Concise and Complete: Multi-stage Training For Adaptive Reasoning

Auteurs: Nathanaël Carraz Rakotonirina, Ren Pang, Neha Anna John, Michael Bohlke-Schneider, Momchil Hardalov
Publié dans: Findings of the Association for Computational Linguistics: ACL 2026, 2026, Page(s) 12798-12810
Éditeur: Association for Computational Linguistics
DOI: 10.18653/v1/2026.findings-acl.622

Can discrete information extraction prompts generalize across language models?

Auteurs: N. Rakotonirina, R. Dessì, F. Petroni, S. Riedel, M. Baroni
Publié dans: Proceedings of ICLR 2023 (International Conference on Learning Representations), 2023
Éditeur: ICLR

Emergence of a high-dimensional abstraction phase in language transformers

Auteurs: Emily Cheng, Diego Doimo, Corentin Kervadec, Iuri Macocco, Jade Yu, Alessandro Laio and Marco Baroni
Publié dans: Proceedings of ICLR 2025, 2025
Éditeur: ICLR

Peer reviewed articles (2)

LiSeCo: Linear Semantic Control for Language Generation

Auteurs: Emily Cheng, Carmen Amo Alonso
Publié dans: Transactions of Machine Learning Research, 2026, ISSN 2835-8856
Éditeur: Journal of Machine Learning Research Inc

Referential communication in heterogeneous communities of pre-trained visual deep networks

Auteurs: Mateo Mahaut, Francesca Franzon, Roberto Dessi, Marco Baroni
Publié dans: Transactions on Machine Learning Research, 2025, ISSN 2835-8856
Éditeur: Journal of Machine Learning Research Inc