Short papers
Up to 4 pages, not including references or appendix.
Machine Learning/CORE Unranked
LCFM @ Neurips 2026
Neurips 2026 Workshop Long Context Foundation Model
The Long Context Foundation Models (LCFM) workshop at NeurIPS 2026 focuses on advancing foundation models capable of effectively utilizing long context windows—beyond 128K to 1M tokens—in agentic frameworks. It emphasizes challenges in long-horizon reasoning, memory, planning, tool use, multimodal interaction, and evaluation, with a focus on applications in scientific discovery and beyond.
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Long papers
Up to 8 pages, not including references or appendix.
Research areas in scope
01
Topics of Interest
Long-context and long-horizon agentic foundation modelsNovel modeling and training approachesEfficiency techniques for foundation modelsRobust evaluationLong-context reasoningLong-context multi-modal learningLong-horizon AI for science
Policies worth checking twice
- Submissions must be anonymized and uploaded as a single PDF to OpenReview.
- References and appendix pages are unlimited but reviewers are not required to read them.
- Posting on preprint servers like ArXiv is permitted.
- Submissions under review at other venues (e.g., ICLR 2027) are allowed, provided they do not violate the other venue's dual-submission or anonymity policy.
- The review process is double-blind.
- This is a non-archival workshop; accepted papers will not be indexed or have archival proceedings.
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026