Call for Papers: KR meets Machine Learning and Explanation

Machine learning (ML) has seen groundbreaking advancements across countless tasks in recent years, not least with the ongoing development and widespread deployment of large language models (LLMs). Concurrently, with this increase in both power and scope, it has become increasingly important that AI models are designed to be supplemented with explanations such that their outputs can be assessed, understood and modified if necessary. Meanwhile, the field of KR presents an excellent repertoire of technologies for leveraging knowledge in both ML and explanation pipelines. This special track thus aims to focus on the synergistic interactions between KR and these complementary fields of ML and explanation.

We welcome contributions that extend the state-of-the-art at the intersection of KR with either of the fields mentioned above. With regards to explanation, contributions that use KR in the explanation of AI models, or that explain numeric and/or symbolic models themselves, are welcomed. From the ML side, these may also include the use of KR methods for solving ML challenges, the use of ML methods for solving KR challenges or the integration of learning and reasoning towards better modelling, solving or explaining in different tasks. Papers focusing on evaluation protocols and benchmarking of these hybrid solutions will be also welcome.

Important Dates

• Submission of title and abstract: February 12, 2026
• Paper submission deadline: February 19, 2026
• Author response period: March 24-28, 2026
• Notification of acceptance: April 13, 2026
• Camera-ready due: May 3, 2026

Topics of Interest

We welcome papers on a wide range of topics where KR is a key component, including (but not limited to):

• Learning symbolic knowledge, such as ontologies and knowledge graphs, action theories, commonsense knowledge, spatial and temporal theories, preference models and causal models
• KR, ML and reasoning in the computation, synthesis, analysis, verification, reuse, and repair of plans
• Logic-based, logical and relational learning algorithms
• ML-driven reasoning algorithms
• Neural-symbolic learning
• Statistical relational learning and KR
• Symbolic reinforcement learning
• The use of KR techniques for supplementing or evaluating LLMs
• LLMs for supporting KR-driven methods
• Knowledge-driven natural language understanding and dialogue
• Learning symbolic abstractions from unstructured data
• Expressive power of learning representations and explanations
• Knowledge-driven decision making and explanations
• Combining discrete and continuous, quantitative and qualitative, logical and probabilistic representations and reasoning methods (e.g., task planning with motion planning) with explanations
• Architectures that combine data-driven techniques and formal reasoning
• KR-driven Explainable AI
• Interpretable ML models intertwined with KR
• Combining KR and ML for enhanced explainability
• Theoretical frameworks for explainability within KR and logic-based systems
• Explainability in dynamic and temporal knowledge representation
• Scalable approaches for real-time explainable reasoning using KR and ML
• Evaluation protocols, metrics, and benchmarks for assessing the quality and clarity of explanations
• Interactive and adaptive explanation frameworks using ML and KR
• Personalisation of explanations through contextual KR, ML and user feedback
• Identifying and mitigating systemic biases in AI explanations through KR
• Hands-on tools and open-source libraries for explanations in real-world settings

Submission Guidelines

Submissions to the special track must fall into the intersection of KR and either ML or explanation, or both. Papers not meeting this criterion will be identified before the actual reviewing process and will be desk-rejected. Generative AI models do not satisfy the criteria for authorship of papers published in KR 2026. If authors use an LLM in any part of the paper-writing process they assume full responsibility for all content, including checking for plagiarism and correctness of the entire submission. Text generated by an LLM as part of the paper’s methodology or experimental analysis is allowed but needs to be properly documented and described in the paper.

Submissions can be either of the following types of technical contributions:

• Long papers (up to 9 pages excluding references and acknowledgements)
• Short papers (up to 4 pages excluding references and acknowledgements)

Both kinds of papers must be prepared and submitted according to the author guidelines on the submission page before the deadline (more information about submissions are available at https://kr.org/KR2026/submission.html). In particular, submissions should be anonymous.

Authors may submit a separate PDF or ZIP file with additional information supporting their claims (such as proof details, additional experimental results, further details on experimental design, etc). The ZIP file (size up to 100 MB) may include content such as PDFs, code, or data. Such supplementary material should be submitted via the conference management system. The paper must be self contained, as the supplementary material will not be published. Reviewers will have the option, but not the obligation, to consult the supplementary material.

Accepted papers will be published in the KR 2026 proceedings. At least one author of each accepted paper is required to participate in the conference and present the work.

Selected authors will be given the option to showcase their work in the Demo Track alongside their regular presentation slot in a session of the track.

Top papers from KR 2026 will be invited to the award-winning paper tracks of Artificial Intelligence (AIJ) and the Journal of Artificial Intelligence Research (JAIR). Thus, award winners will have the possibility of choosing between AIJ and JAIR.

Inquiries

Inquiries should be sent by email to kr26-mlx-chairs@lirmm.fr and will be handled by the KR meets Machine Learning and Explanation Track chairs:

• Claudia d’Amato, University of Bari, Italy
• Ute Schmid, University of Bamberg, Germany
• Antonio Rago, King’s College London, UK