Autonomous Linguistic Emergence in neural Networks

Información del proyecto

ALiEN

Identificador del acuerdo de subvención: 101019291

DOI

10.3030/101019291

Fecha de la firma de la CE 26 Mayo 2021

Fecha de inicio 1 Enero 2022

Fecha de finalización 31 Diciembre 2026

Financiado con arreglo a

• EXCELLENT SCIENCE - European Research Council (ERC)

Coste total

€ 2 489 541,00

Aportación de la UE

€ 2 489 541,00

Coordinado por

UNIVERSIDAD POMPEU FABRA
Spain

CORDIS proporciona enlaces a los documentos públicos y las publicaciones de los proyectos de los programas marco HORIZONTE.

Los enlaces a los documentos y las publicaciones de los proyectos del Séptimo Programa Marco, así como los enlaces a algunos tipos de resultados específicos, como conjuntos de datos y «software», se obtienen dinámicamente de OpenAIRE .

Publicaciones

Conference proceedings (19)

The emergence of numerical representations in communicating artificial agents

Autores: Daniela Mihai, Lucas Weber, Francesca Franzon
Publicado en: Proceedings of Evolang XVI, 2026, Página(s) 308-315
Editor: The Evolution of Language Conferences

Cross-domain image captioning with discriminative finetuning.

Autores: R. Dessì, M. Bevilacqua, E. Gualdoni, N. Rakotonirina, F. Franzon, M. Baroni
Publicado en: Proceedings of CVPR 2023 (IEEE/CVF Conference on Computer Vision and Pattern Recognition), 2023, Página(s) 6935-6944
Editor: Computer Vision Foundation

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

Autores: Jacobs, C., Rakotonirina, N.C., Chimoto, E.A., Bassett, B.A., Kamper, H.
Publicado en: Proc. INTERSPEECH 2023, 2023
Editor: ISCA
DOI: 10.21437/interspeech.2023-421

On the Correspondence between Compositionality and Imitation in Emergent Neural Communication

Autores: E. Cheng, M. Rita, T. Poibeau
Publicado en: Findings of the Association for Computational Linguistics: ACL 2023, 2023, Página(s) 12432–12447
Editor: ACL

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

Autores: M. Mahaut, F. Franzon, R. Dessì, M. Baroni
Publicado en: Proc. of the 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2023), 2023, Página(s) 2619-2621
Editor: International Foundation for Autonomous Agents and Multiagent Systems

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

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

Geometric Signatures of Compositionality Across a Language Model’s Lifetime

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

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

Autores: R. Dessì, E. Gualdoni, F. Franzon, G. Boleda, M. Baroni
Publicado en: Proceedings of EMNLP 2022 (Conference on Empirical Methods in Natural Language Processing), 2022, Página(s) 7998-8007
Editor: ACL

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

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

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

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

Bridging information-theoretic and geometric compression in language models

Autores: Cheng, Emily; Kervadec, Corentin; Baroni, Marco
Publicado en: Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, 2023
Editor: Association for Computational Linguistics

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

Autores: Rakotonirina, Nathanael; Baroni, Marco
Publicado en: Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), 2024
Editor: Association for Computational Linguistics

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

Autores: Richard Antonello, Emily Cheng
Publicado en: Proceedings of UniReps@NeurIPS 2024, 2024
Editor: OpenReview/Online

Prediction Hubs are Context-Informed Frequent Tokens in LLMs

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

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

Autores: Kervadec, Corentin; Franzon, Francesca; Baroni, Marco
Publicado en: Findings of EMNLP, 2023
Editor: Association for Computational Linguistics
DOI: 10.18653/v1/2023.findings-emnlp.959

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

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

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

Autores: Nathanaël Carraz Rakotonirina, Ren Pang, Neha Anna John, Michael Bohlke-Schneider, Momchil Hardalov
Publicado en: Findings of the Association for Computational Linguistics: ACL 2026, 2026, Página(s) 12798-12810
Editor: Association for Computational Linguistics
DOI: 10.18653/v1/2026.findings-acl.622

Can discrete information extraction prompts generalize across language models?

Autores: N. Rakotonirina, R. Dessì, F. Petroni, S. Riedel, M. Baroni
Publicado en: Proceedings of ICLR 2023 (International Conference on Learning Representations), 2023
Editor: ICLR

Emergence of a high-dimensional abstraction phase in language transformers

Autores: Emily Cheng, Diego Doimo, Corentin Kervadec, Iuri Macocco, Jade Yu, Alessandro Laio and Marco Baroni
Publicado en: Proceedings of ICLR 2025, 2025
Editor: ICLR

Peer reviewed articles (2)

LiSeCo: Linear Semantic Control for Language Generation

Autores: Emily Cheng, Carmen Amo Alonso
Publicado en: Transactions of Machine Learning Research, 2026, ISSN 2835-8856
Editor: Journal of Machine Learning Research Inc

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

Autores: Mateo Mahaut, Francesca Franzon, Roberto Dessi, Marco Baroni
Publicado en: Transactions on Machine Learning Research, 2025, ISSN 2835-8856
Editor: Journal of Machine Learning Research Inc