CoNLL 2025 | CoNLL

Vienna, Austria, July 31 - August 1, 2025 (co-located with ACL)

Welcome to the 29th Conference on Computational Natural Language Learning (CoNLL 2025). CoNLL is a yearly conference organized by SIGNLL (ACL's Special Interest Group on Natural Language Learning), focusing on theoretically, cognitively, and scientifically motivated approaches to computational linguistics. In particular, it seeks to explore the interaction between theoretical issues in linguistics and cognition, on the one hand, and computational modeling, on the other. Questions that CoNLL addresses are e.g.: How do computational models inform us about how language and cognition work? Which phenomena do they capture, which do they not capture? Conversely, how do linguistics and cognitive science contribute to understanding, and possibly improving, AI models?

We welcome work targeting any aspect of language and its computational modeling, including but not only:

Computational Psycholinguistics, Cognition and Linguistics
Computational Social Science and Sociolinguistics
Interaction and Dialogue
Language Acquisition, Learning, Emergence, and Evolution
Multimodality and Grounding
Typology and Multilinguality
Speech and Phonology
Syntax and Morphology
Semantics and Pragmatics
Theoretical Analysis and Interpretation of ML Models for NLP
Resources and Tools for Scientifically Motivated Research

Invited Speakers

Jean-Rémi King, CNRS / Meta AI
Title: Emergence of Language in the Human Brain
Abstract:
Deep learning has made major progress in natural language processing. Beyond these technical performance, these algorithms offer new methods to understand and model how language is processed in the human brain. Using both encoding (representation -> brain) and decoding (brain -> representations), we show that the comparison between modern speech and language models effectively accounts for brain responses to natural speech as recorded with EEG, MEG, iEEG and fMRI, including in children between 2 and 12 years old. This systematic comparison provides an operational foundation to model language in the adult and developing brain, and thus offers a new path to understand the neural and computational bases of this human-specific ability.

Raquel Fernández, University of Amsterdam
Title: The Interplay between Gestures and Speech in Face-to-Face Communication
Abstract:
The primary form of language use is face-to-face dialogue — an inherently multimodal setup where we use speech in tandem with non-verbal signals such as gaze and gestures. However, how speech and gestures interact in face-to-face communication is not yet well understood. I will present our work on gesture representation learning, which aims to shed light on this question. We propose a self-supervised learning approach that grounds body movements in spoken language to learn representations of iconic gestures in referential tasks. Our experiments show that the resulting gesture embeddings exhibit properties that align with theoretically motivated hypotheses and highlight the complementary roles of gesture and speech in face-to-face reference resolution. More generally, our work advances the use of data-driven computational modelling to study fundamental properties of multimodal communication.

Author instructions:

Oral presentations should be 15 minutes in length. A common laptop will be provided. To ensure a smooth experience, we collect presentation slides in advance via Underline and upload them to the designated laptops.

Poster presentations should be A0 in portrait format. If you need local printing in Vienna, please check the ACL 2025 FAQ: https://2025.aclweb.org/faq/#Print-Shops-for-Poster

Online presentation (optional): We kindly invite you to submit a pre-recorded video presentation (max. 15 minutes) and/or poster via Underline by July 10, 2025. The submission form can be found here: https://acl2025-workshops.paperform.co/ (select W:15 29th Conference on Computational Natural Language Learning).

Detailed instructions on preparing and uploading your materials to Underline are available here: https://docs.google.com/presentation/d/1B26aSIWCRTYEFKvndrYWspNBUpimN6v…

Timeline
(All deadlines are 11:59pm UTC-12h, AoE)

~~Submission deadline (archival and non-archival): March 14 2025~~
~~ARR Commitment deadline CoNLL only accepts direct submissions this year!~~
~~Notification of acceptance: May 23 2025~~
~~Camera-ready papers due: June 17 2025~~
Conference: July 31 - August 1, 2025

Venue/Visa

CoNLL 2025 will be held in Vienna, Austria. Please note that it is an in-person conference. We expect all accepted papers to be presented physically and presenting authors must register through ACL (workshop). If you need a visa, we encourage you to submit your visa application at submission time. See here for more information.

Contact

Questions? E-mail conll.chairs@gmail.com

Social Media

Program Chairs

Gemma Boleda, Universitat Pompeu Fabra / ICREA
Michael Roth, University of Technology Nuremberg

Area Chairs

Christian Bentz, University of Passau
Zhenguang G. Cai, Chinese University of Hong Kong
Tanise Ceron, Bocconi University
Jackie Chi Kit Cheung, McGill University
Iria De Dios, Universitat Pompeu Fabra
Max van Duijn, Leiden University
Kilian Evang, HHU Düsseldorf
Agnieszka Faleńska, University of Stuttgart
Aina Garí Soler, INRIA
Ximena Gutierrez-Vasques, Universidad Nacional Autónoma de México
Dieuwke Hupkes, Meta AI Research
Xixian Liao, Barcelona Supercomputing Center
Brielen Madureira, University of Potsdam
Yohei Oseki, University of Tokyo
Sandro Pezzelle, University of Amsterdam
Emily Prud'hommeaux, Boston College
Tanja Samardžić, University of Zurich
Carina Silberer, University of Stuttgart
Ece Takmaz, Utrecht University
Tessa Verhoef, Leiden University
Yang Xu, University of Toronto
Wei Zhao, University of Aberdeen

Publicity Chairs

Snigdha Chaturvedi, University of North Carolina at Chapel Hill
Anvesh Rao Vijjini, University of North Carolina at Chapel Hill

Publication Chairs

Emily Cheng, Universitat Pompeu Fabra
Selina Meyer, University of Technology Nuremberg

Platinum Sponsor