Poster Papers
Accepted papers are presented as posters during poster sessions.
Machine Learning/CORE Unranked
Sci4DL 2026
Workshop on Scientific Methods for Understanding Deep Learning
The Workshop on Scientific Methods for Understanding Deep Learning (SciForDL 2026) is a venue for researchers to apply the scientific method to investigate how and why deep learning systems work. It emphasizes empirical experiments, hypothesis testing, and reproducible studies over traditional criteria like state-of-the-art performance or rigorous theorems, aiming to bridge theory and practice in deep learning research.
Key deadlines
Verified September 9, 2026Abstract registration
February 5, 2026
AoE
Full paper
February 5, 2026
AoE
Workshop timeline
Submission and decisions
Abstract registrationKey deadline
February 5, 2026 · AoE
Full paperKey deadline
February 5, 2026 · AoE
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Contributed Talks
Selected works are highlighted as contributed talks.
Challenge Submission
Authors can opt-in to the Science of DL Improvement Challenge for a chance to present a lightning talk.
Research areas in scope
01
Application Domains
visionlanguage(self)-supervised learninggenerative models
02
Network Architectures
CNNstransformers
03
Objects of Study
training dynamicsloss landscapesweightsrepresentations
Policies worth checking twice
- Submissions are limited to 4 pages, with unlimited additional pages for references and appendices.
- The review process is double-blind; all submissions must be anonymized.
- Submissions that breach anonymity will be desk-rejected.
- If a submission was previously accepted at a conference, journal, or workshop (including ICLR 2026), it must be extended with new results to be considered.
- Papers currently under review elsewhere are welcome.
- There will be no formal proceedings; accepted submissions will be made public on OpenReview.
- Submissions must be made via OpenReview.
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026