Interdisciplinary Institute in Artificial Intelligence

Lorraine Goeuriot,
Associate professor,
lorraine.goeuriot@univ-grenoble-alpes.fr







François Portet,
Professor,
francois.portet@univ-grenoble-alpes.fr

DESCRIPTION

The AI and Language Chair aims to design and develop AI models capable of processing natural language in different modalities (text, voice, sign language, images) for the benefit of humanity.

These models should:
1) be adapted to specialized languages;
2) be explainable and free from social bias;
3) be able to capture stable and coherent meaning in language; 4) be able to leverage external knowledge
5) be used by experts and non-experts without harm and for the benefit of humanity.

While large language models (LLMs) and Retrieval Augmented Generation (RAG) have achieved remarkable breakthroughs, the aforementioned objectives remain challenging. To address these challenges, the chair brings together tenured faculty, Ph.D. students, and master’s students, who will receive direct support from the chair. Furthermore, the chair’s activities will be carried out in direct collaboration with companies.

Structure of the chair:

Work Package 1: NLP models for specialized domains
Work Package 2: Evaluation and Interpretability of NLP models
Work Package 3: Grounding LLMs in the physical world
Work Package 4: Leveraging Knowledge for Model Enhancement
Work Package 5: NLP and machine-generated content

ACTIVITIES

SELECTED LIST OF PUBLICATIONS

• Méloux, M., Maniu, S., Portet, F., & Peyrard, M. (2025) “Everything, Everywhere, All at Once: Is Mechanistic Interpretability Identifiable?” In The Thirteenth International Conference on Learning Representations.
• Herron, F., Rossato, S., Allauzen, A., Favre, B., & Portet, F. (August 2025). Speaker Group Encoding in Self-supervised Speech Recognition Models. In International Conference on Text, Speech, and Dialogue (pp. 121–132). Cham: Springer Nature Switzerland.
• Cancellieri, M., El-Ebshihy, A., Fink, T., Galuščáková, P., Gonzalez-Saez, G., Goeuriot, L., ... & Schaer, P. (April 2025). LongEval at CLEF 2025: Longitudinal Evaluation of IR Model Performance. In European Conference on Information Retrieval (pp. 382–388). Cham: Springer Nature Switzerland.
• Parcollet, T., Nguyen, H., Evain, S., Boito, M. Z., Pupier, A., Mdhaffar, S., ... & Besacier, L. (2024). Lebenchmark 2.0: A standardized, replicable, and enhanced framework for self-supervised representations of French speech. Computer Speech & Language, 86, 101622.
• Segonne, V., Mannion, A., Canul, L. C. A., Audibert, A., Liu, X., Macaire, C., ... & Schwab, D. (May 2024). Jargon: A suite of language models and evaluation tasks for specialized domains in French. In The 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024) (pp. 9463–9476).
• Desot, T., Portet, F., & Vacher, M. (2022). End-to-End Spoken Language Understanding: Performance Analyses of a Voice Command Task in a Low-Resource Setting. Computer Speech & Language, 75, 101369.
• Evain, S., Nguyen, M. H., Le, H., Boito, M. Z., Mdhaffar, S., Alisamir, S., ... & Besacier, L. (December 2021). Task-agnostic and task-specific self-supervised learning from speech using Lebenchmark. In the 35th Conference on Neural Information Processing Systems (NeurIPS 2021).
• G. Cabanac, C. LabbÉ, and A. Magazinov. “Tortured Phrases: A Dubious Writing Style Emerging in Science. Evidence of Critical Issues Affecting Established Journals.” arXiv (2021). https://doi.org/10.48550/arXiv.2107.06751

CHAIR PRESENTATION

License: