The 10th Workshop on Representation Learning for NLP (RepL4NLP 2025), co-located with NAACL 2025 in Albuquerque, New Mexico, invites papers of a theoretical or experimental nature describing recent advances in vector space models of meaning, compositionality, and the application of deep neural networks and spectral methods to NLP. We welcome submissions on representations of text, as well as representations that are multi-modal, cross-lingual, representations of symbolic languages, code, enriched with external knowledge, or structure-informed (syntax, morphology, etc). Topics for the workshop will include, but are not limited to:
Efficient learning of representations and inference as models scale up: with respect to the amount of training and fine-tuning data, training and inference computing time, and model energy consumption for development and deployment.
Investigating the representation dynamics during training: understanding how representations evolve throughout the training process
Evaluating existing representations: probing representations for generalization, compositionality, robustness, etc.
Understanding the relationship between representations and model behaviors: how learned representations drive predictions, how interventions in the representation space causally affect the model behaviors, and what types of representations lead to better simulation of human behaviors, etc.
Beyond English textual representation: including but not limited to cross-modal, cross-lingual, knowledge-informed, linguistically-informed, and cognitively plausible representations, and how data from different sources interact in the training and inference processes.
Developing new representations: at various levels, using language model objectives, spectral methods, neuro-symbolic methods, etc.
Direct paper submission deadline: January 30, 2025
ARR commitment deadline: February 20, 2025
Notification of acceptance: March 1, 2025
Camera-ready paper due: March 10, 2025
Workshop date: May 4, 2025
We solicit three categories of papers: regular workshop papers, extended abstracts and cross-submissions. Only regular workshop papers will be included in the proceedings as archival publications, and only regular workshop papers will be eligible for best paper prizes. All categories of papers may be long (maximum 8 pages of content + up to one page for limitations (required) + unlimited references) or short (maximum 4 pages of content + up to one page for limitations (required) + unlimited references). Please refer to NAACL website for more detail about the formats. We encourage authors to include a broader impact and ethical concerns statement, following ARR Ethics Policy from the main conference. Authors can submit their papers directly via OpenReview at: https://openreview.net/group?id=aclweb.org/NAACL/2025/Workshop/RepL4NLP. Alternatively, submissions through ARR can be committed to the workshop using the following link: https://openreview.net/groupid=aclweb.org/NAACL/2025/Workshop/RepL4NLP_ARR_Commitment.
Please note that this year’s accepted papers will not require a video submission.
Preliminary but interesting ideas or results that have not been published before may be submitted as extended abstracts. Accepted extended abstracts will be presented as posters, and included in the workshop program and handbook, but will not be included in the workshop proceedings (and are not eligible for best paper prizes).
Extended abstract submissions are therefore ideal for preliminary work which would benefit from exposure but is not ready for publication. Authors should submit a long extended abstract of up to 8 pages, with up to 2 additional pages for references, or a short extended abstract of up to 4 pages, with up to 2 additional pages for references. Submissions should follow the NAACL 2025 formatting requirements (see the NAACL 2025 Call For Papers for reference: https://2025.naacl.org/calls/papers/). Reviewing will be double-blind, and thus no author information should be included in the papers; self-reference that identifies the authors should be avoided or anonymized.
In addition to previously unpublished work, we also solicit papers on relevant topics that have appeared in a non-NLP venue (e.g., workshop or conference papers at NeurIPS or ICML) or findings papers in NLP venues. Accepted cross-submissions will be presented as posters, with an indication of original venue, but will not be included in the workshop program and handbook, or the workshop proceedings (and are not eligible for best paper prizes). Cross-submissions are ideal for related work which would benefit from exposure to the RepL4NLP audience. Submission length is determined by the original venue. Interested authors should submit their papers in PDF format through the RepL4NLP submission portal with a note on the original venue. Papers in this category do not need to follow the ACL format and selection will be solely determined by the organising committee. These submissions do not undergo the standard reviewing process and the organizing comitee makes the decision on whether they will be presented. We ask the authors to submit these works directly via email to repl4nlp@googlegroups.com.
Contact Information
The co-chairs of the workshop can be contacted by email at: repl4nlp@googlegroups.com.