We invite researchers from machine learning and related fields to submit their latest work on theory and applications of efficient spatial reasoning to the workshop. We encourage submissions related (but not limited) to the following topics:
Novel architectures for efficient spatial reasoning, including sparse, modular, and geometry-aware designs.
Training and optimization techniques such as distillation, token-budgeting, curriculum learning, and reinforcement learning for concise reasoning.
Benchmarking and evaluation methods that jointly assess accuracy, efficiency, interpretability, and robustness in spatial reasoning tasks.
Applications of efficient spatial reasoning in domains such as robotics, AR/VR, scientific discovery, urban planning, and simulation.
Safety, fairness, and ethical considerations when compressing or pruning spatial reasoning processes.
Submission deadline: 2026/02/07, 11:59PM AOE;
Acceptance notification: 2026/03/01
Formatting: All submissions must be PDFs in the ICLR format. We accept Regular (9 pages) and Tiny (4 pages) papers, with unlimited pages for references and supplementary materials. Code may be included as a supplementary zip or via an anonymized repository link. Regular papers have a higher likelihood of being selected for an oral presentation.
Reviewing: Reviewing will be double-blind, so authors should take care to adequately anonymize their submission, especially regarding supplementary code. Reviewers will be instructed to focus on correctness and relevance to the workshop.
Non-archival: This workshop is non-archival and will not result in proceedings; workshop submissions can be submitted to other venues.
Dual submission: We welcome papers that may have been already accepted at ICLR 2026 but which are also relevant to this workshop, or that under reviews at other venues (e.g., ICML 2026).
Contact: Reach out to esreasoning.workshop@gmail.com