Laboratoire d'Informatique de Paris Nord

🚀 Excited to share that a survey paper from our RCLN team has been accepted at IJCAI 2026! This work has been done in collaboration with CentraleSupélec.
"Graph Rewiring in GNNs to Mitigate Over-Squashing and Over-Smoothing: A Survey" By Hugo Attali, Nathalie Pernelle, Davide Buscaldi, and Fragkiskos D. Malliaros.

📎 https://arxiv.org/abs/2411.17429

Graph Neural Networks are powerful models for learning from graph-structured data, yet their effectiveness is often limited by two critical challenges: over-squashing, where information from distant nodes is excessively compressed, and over-smoothing, where repeated propagation makes node representations indistinguishable. Both phenomena stem from the interaction between message passing and the input topology, ultimately degrading information flow and limiting the performance of GNNs.Our survey also opens a broader discussion on the limits and open questions of graph rewiring: when is modifying the topology truly necessary? How can observed improvements be properly attributed to connectivity changes rather than feature-driven effects? We argue that progress in this area will require clearer problem formulations, more explicit assumptions, and evaluation protocols that make results robust and comparable across settings positioning graph rewiring as a principled structural intervention to better understand how topology shapes learning in GNNs.

Looking forward to presenting our work at IJCAI 2026 at Bremen!

#GNN #MachineLearning #IJCAI2026 #GraphNeuralNetworks #AI #DeepLearning #Research #LIPN

Proud to share that the RCLN team of the LIPN laboratory will be presenting two papers at premier IR and NLP conferences this July! 🏆

1️⃣ SIGIR 2026 : Voronoi Token Pruning in Late-Interaction Models

By Yash Kankanampati, Yuxuan Zong, Nadi Tomeh, Benjamin Piwowarski, and Joseph Le Roux.

🔗 https://arxiv.org/abs/2603.09933

2️⃣ ACL 2026 : Emergent Mention Detection in LLMs — Accepted at #ACL2026

By Victor Morand, Nadi Tomeh, Josiane Mothe, and Benjamin Piwowarski.

🔗 https://arxiv.org/abs/2510.19410

We look forward to an exciting month of July sharing these advancements!

🎉Thrilled to announce that the paper "PUMA: Projected Universal Multilingual ASR for Low-Resource Settings" has been accepted at the #ACL2026 Findings Conference!
Congratulations to Ilyes Oukid, Bilal Faye, Hanane Azzag, Mustapha Lebbah and Said Yacine Boulahia, all proud members of the Laboratoire d'Informatique de Paris-Nord (LIPN).
This work was carried out in collaboration with the DAVID Laboratory at University Versailles Saint-Quentin-en-Yvelines/ Université Paris-Saclay.

See you in San Diego, USA!

Two teams from LIPN will present their joined work at IPMU 2026 👏.
Congratulations to Amal Beldi and Louenas Bounia for their work on Uncertainty-Aware Contextual Recommendation under Possible Worlds Semantics!
This paper proposes a probabilistic framework for uncertainty-aware contextual recommendation grounded in probabilistic database semantics.
#LIPN #RecommenderSystems #DecisionMaking

👏 Congratulations to Jaime Arias Almeida, a CNRS Research Engineer @LipnLab, who is now in charge of Software at the CNRS Sciences Informatiques:
🔗 https://www.ins2i.cnrs.fr/fr/personne/jaime-arias-almeida

❓ Do we really know what an algorithm is? To learn more about interdisciplinary work between philosophy and computer science, check out the CNRS online journal:
🔗 https://lejournal.cnrs.fr/nos-blogs/focus-sciences/algorithme-un-mot-simple-plein-dambiguites
🤝 This is part of a joint project between the LIPN and the IHPST (Paris), carried out by T. Seiller and A. Naibo, funded by the CNRS and the ANR.

We are pleased to announce a new publication, a long paper at #CIKM2025 (Seoul, Korea), from the A3 and RCLN teams at the LIPN laboratory, in collaboration with Centrale Supélec. This paper, titled “Dynamic Triangulation-Based Graph Rewiring for Graph Neural Networks” proposes a new graph rewiring method that dynamically selects task-relevant triangles to mitigate oversquashing in GNNs. Congratulations to Hugo Attali (LIPN), Thomas Papastergiou (LIPN), Nathalie Pernelle (LIPN) and Fragkiskos Malliaros (Centrale Supélec) ! 👏

💫 Congratulations to Aude Grezka @grezka, a CNRS Research Engineer at @LipnLab and one of the new ambassadors for the "La Science taille XX elles" program!
🔗 https://www.paris-centre.cnrs.fr/fr/cnrsinfo/la-science-taille-xx-elles-lexposition-qui-celebre-les-femmes-scientifiques-revient-pour

📆 Today is the day for the MathStic workshop on categories, involving both the @LipnLab and the neighbouring LAGA mathematics laboratory.
👩‍🏫 Speakers are Jad Koleilat, Samuel Mimram, Paula Verdugo and Sacha Ikonicoff.
▶️ https://lipn.univ-paris13.fr/~breuvart/journee_mathstic_axe_categories

☀️ Three papers from the Local team of the LIPN lab were accepted to the FSCD 2025 conference :
https://fscd2025.github.io/
❓ These papers study the categorical semantics of Concurrency, Complexity or Differentiation, confirming the strong categorical expertise of the Logic team at the @LipnLab!
🤝 Congratulations to Flavien Breuvart and Hugo Paquet (now Inria), Baptiste Chanus, Damiano Mazza and Morgan Rogers, and Marie Kerjean, Valentin Maestracci and Morgan Rogers.

👏 Proud to share a new paper accepted in ACL 2025 from IRISA and LIPN laboratories: Bregman CRFs for sequence labelling! Congratulations to Caio Corro (IRISA), Mathieu Lacroix (LIPN) and Joseph Leroux (LIPN)!
🔗 https://caio-corro.fr/pdf/bregman_crf_acl_2025.pdf

💫 The @LipnLab is thrilled and proud to welcome Meena Mahajan for one month as an invited professor, starting May 19th.
🎓 Pr. Mahajan is a renowned international expert in complexity theory, notably algebraic complexity. She will be hosted at the @LipnLab by Sylvain Perifel and Pascal Weil.
🔗 https://www.imsc.res.in/~meena/

⚕️ New Dataset on French Medical Dispatch! We are thrilled to announce that the new SIMAMU dataset dedicated to French Medical Dispatch Dialog is accepted to the Computer Methods and Programs in Biomedicine journal. This work results from a collaboration between Inserm, Inria HeKA, AP-HP Department of Medical Informatics, and LORIA and LIPN laboratories.
🎓 Congratulations to Aimé Nun (Inserm, Inria, AP-HP), Olivier Birot (Inserm, Inria), Gaël Guibon (Sorbonne Paris Nord, LIPN, Université de Lorraine, LORIA), Frédéric Lapostolle (SAMU 93, AP-HP, Sorbonne Paris Nord) and Ivan Lerner (Inria, AP-HP) for this achievement!
🔗 You can already access the dataset on Huggingface: https://huggingface.co/datasets/medkit/simsamu