Full paper
Original research or position papers not exceeding 10 pages, excluding references and supplementary materials; submitted via OpenReview.
Machine Learning/CORE Unranked
NeSy 2026
20th International Conference on Neurosymbolic Learning and Reasoning
NeSy 2026 is the 20th edition of the premier international conference on neurosymbolic learning and reasoning, organized annually since 2005 by the Neurosymbolic Learning and Reasoning Association. The conference brings together researchers and practitioners to present cutting-edge work integrating neural and symbolic AI, with applications across healthcare, finance, robotics, and more, and features both academic and industry tracks.
Paper fit
Contribution paths
A strong submission should clearly identify its contribution and evaluate it appropriately.
Short paper
Original research or position papers not exceeding 5 pages, excluding references and supplementary materials; submitted via OpenReview.
PhD symposium paper
Overview of PhD research plan in neurosymbolic AI, not exceeding 4 pages; submitted via OpenReview.
Industry paper
Overview of industrial effort related to neurosymbolic AI, 3 to 4 pages excluding references; submitted via EasyChair.
Research areas in scope
01
Main Track Subject Areas
Informed or physics-aware Machine LearningAddressing knowledge representation and reasoning tasks using neural networksCode generation and knowledge engineering with neural networks, including with LLMsStudying and improving LLM reasoning with Neurosymbolic methodsNeurosymbolic cognitive modellingLanguages for Neurosymbolic AIEmbedding methods for structured informationCircuits and knowledge compilation for Neurosymbolic AISpecification and verification of machine/deep learning systemsNeurosymbolic AI frameworksNeurosymbolic methods for reinforcement, causal, structure, transfer, meta, multitask, continual, or relational - learning, and graph neural networksApplications of Neurosymbolic AI
02
Special tracks
Neurosymbolic Generative ModelsKnowledge Graphs, Ontologies and Neurosymbolic AINeurosymbolic Methods for Trustworthy and Interpretable AI
03
Industry Track: Topics of Interest
Case studies and success stories of deployed neuro-symbolic systems in specific verticals, such as finance, healthcare, robotics, or manufacturingBest practices and lessons learned from integrating neural and symbolic methods in an industrial contextIntegration of Large Language Models (LLMs) with symbolic reasoning to improve factuality, planning, and reliabilityDiscussions on scalability, reliability, and other challenges encountered in production environmentsDesign and implementation of guardlines for deep learning systems using symbolic methods, such as mathematical reasoningReports on barriers to adoption and strategies for overcoming themBenchmarks, datasets, tasks, and evaluation metrics or approaches with an industry contextAnalyses of business value, ROI, and competitive advantages gained from using hybrid AI
Policies worth checking twice
- All submissions must be anonymized and must not compromise the double-blind review process.
- Double submissions are strictly prohibited.
- Full papers must not exceed 10 pages; short papers and late submissions must not exceed 5 pages; PhD and industry papers must not exceed 4 pages.
- Authors may share versions of their work on preprint platforms such as arXiv.
- At least one author of each accepted paper must physically attend the conference to present.
- First-round resubmissions must include a changelog and detailed response to reviewers' comments and will be reviewed by the same reviewers if possible.
- An 'Reject and Resubmit' (R&R) decision is offered, which does not guarantee acceptance in the second round.
- Accepted papers will be published in Proceedings of Machine Learning Research (PMLR).
Official sources
Compiled from the official call for papers. The organizers’ pages remain authoritative.
Last verified September 9, 2026