EALM 2025
Ethic and Alignment of (Large) Language Models

30 June 2025
CORIA-TALN
Aix-Marseille Université - Site de Saint-Charles, Marseille

EALM 2025@CORIA-TALN

The EALM 2024 Workshop, organized jointly with the CORIA-TALN 2025 conferences, aims to explore the ethics and alignment of language models.

The workshop will take place on June 30, 2025, as a half-day event in Marseille and will feature several keynote talks as well as a poster session.

Workshop Topics

  • Language Model Alignment
  • Fairness in Language Models (datasets and methods)
  • Ethical Considerations in Language Model Development
  • Efficient Training and Deployment of Language Models
  • Societal Impacts of Language Models
  • Intercultural Considerations
  • Human Preference Elicitation
  • Inclusive Language Models

Call for Contributions

As part of the workshop, we are launching a call for contributions for the poster session, offering an opportunity to present work related to the workshop’s topics.

A subset of selected submissions will be invited for oral presentations.

Submission Details

  • Format:
    2-page position paper in English (with unlimited pages for appendix and references)
  • Submission: Submission Link
    Please select the workshop in the second step when creating a new submission.
  • Formatting Guidelines:
    Submissions should adhere to the main conference formatting guidelines, available here: Template for CORIA-TALN 2025

Important Dates

  • Submission Portal Opens: April 15, 2025
  • Submission Deadline: May 2, 2025
  • Notification of Acceptance: May 9, 2025
  • Final Version Submission: May 16, 2025
  • Workshop Date: June 30, 2025

All deadlines are set to 23:59 UTC -12h (Anywhere on Earth time zone).

Schedule

TimeDescription
09:00 – 09:10Welcome & Introduction
09:10 – 10:10Keynote
Modern Multilingual NLP for Proactive Harmful Content Moderation — Daryna Dementieva slides
10:10 – 10:30Short Presentations Part 1
Comment mesurer les biais politiques des grands modèles de langue multilingues? — Paul Lerner, Laurène Cave, Hal Daumé, Léo Labat, Gaël Lejeune, Pierre-Antoine Lequeu, Benjamin Piwowarski, Nazanin Shafiabadi, François Yvon slides
10:30 – 11:00Coffee Break
11:00 – 11:40Short Presentations Part 2
- MascuLead : le premier tableau de bord de biais de genre — Fanny Ducel, Jeffrey André, Aurélie Névéol, Karën Fort slides
- La Boussole Cassée de l’Alignement Politique — Noé Durandard slides
11:40 – 12:30ANR Diké Feedback
Hubert Curien Laboratory contributions — Thibaud Leteno slides
ERIC Laboratory contributions — Irina Proskurina slides
Naver Labs Europe contributions — Vassilina Nikoulina slides
12:30 – 14:00Lunch Break

Location and Venue




Campus Saint-Charles, Aix-Marseille Université, Marseille

30 June 2025 (09:00–12:30)

Organizing Committee

The workshop is organized by members of the ANR Diké project team:

  • Antoine Gourru ✉, Université Jean Monnet, Télécom Saint-Étienne, Laboratoire Hubert Curien
  • Julien Velcin, Université de Lyon, Lyon 2, Laboratoire ERIC
  • Caroline Brun, Naver Labs Europe
  • Vassilina Nikoulina, Naver Labs Europe
  • Thibaud Leteno ✉ , Université Jean Monnet, Laboratoire Hubert Curien
  • Irina Proskurina ✉, Université de Lyon, Lyon 2, Laboratoire ERIC