The 38th International Conference on Algorithmic Learning Theory (ALT 2027) will be held March 9–12, 2027, in Leiden, the Netherlands. The conference is dedicated to all theoretical and algorithmic aspects of machine learning. We invite submissions on new or existing learning problems, including but not limited to the following topics:
- Design and analysis of learning algorithms
- Classical foundations of learning theory, including statistical, computational, algorithmic, and information-theoretic foundations
- Online learning, multi-agent learning, and game theory
- Optimization, including convex and nonconvex methods, implicit bias, and overparameterization
- Learning paradigms, including supervised, unsupervised, semi-supervised, active, and reinforcement learning
- Reinforcement learning, including classical control-theoretic perspectives, modern applications such as LLM post-training, and new algorithms
- Large language models, transformers, and related theoretical questions
- Theoretical perspectives on trustworthy AI and AI safety, including privacy, adaptive data analysis, fairness, and alignment
- Robustness, including training-data corruption, adversarial examples, and LLM jailbreaks
- Learning under distribution shift, including domain adaptation and out-of-distribution generalization
- Theoretical perspectives on deep learning, including approximation, generalization, and optimization for classical and modern architectures
- Statistics, including asymptotics, high-dimensional statistics, nonparametric methods, and causality
- Learning with algebraic or combinatorial structure
- Bayesian methods
- Kernel methods
- Interpretability and explainability
- Learning under algorithmic constraints, including distributed, communication-efficient, memory-efficient, federated, and streaming learning
- Learning with different data modalities, including time series, sequence-to-sequence mappings, and graph data
- Mathematical analysis of sampling methods, including diffusion models and other practical methods
Despite the theoretical focus of the conference, authors are welcome to support their analysis with relevant empirical results. Accepted papers will be presented at the conference as full-length talks and published electronically in the Proceedings of Machine Learning Research (PMLR).
Important Dates
- Paper submission deadline: October 12, 2026, 11:59 p.m. AoE
- Author feedback: November 30–December 6, 2026
- Author notification: January 5, 2027
Conference Format
The conference will be held in person and will not support remote talks. At least one author of each accepted paper must present the paper in person at the conference. Exceptional circumstances, including visa issues, travel-safety concerns, and medical issues, will be considered on a case-by-case basis.
Publication
Authors of accepted papers may opt out of the proceedings and instead publish a one-page extended abstract that points to an open-access archival version of the full paper reviewed for ALT.
Dual-Submission Policy
Conferences: In general, submissions that are substantially similar to papers previously published, accepted for publication, or submitted in parallel to other peer-reviewed conferences with proceedings may not be submitted to ALT.
Journals: Submissions that are substantially similar to papers already published in a journal at the time of submission may not be submitted to ALT.
Although reviewers will not see author identities, authors may—and are encouraged to—post their papers on arXiv.
Policy on Generative AI
Authors may use generative AI tools in preparing their submissions, but they remain fully responsible for all submitted content, including its correctness, originality, attribution, and citations. Generative AI tools may not be listed as authors.
Authors are strongly encouraged to disclose any substantive use of generative AI in a clearly labeled section titled AI Disclosure, identifying the tool and briefly explaining how it was used. Substantive use may include generating text or mathematical arguments. When AI is used to identify ideas or references, authors are strongly encouraged to consult the original sources. Routine assistance with spelling, grammar, formatting, or identifying potentially relevant references does not need to be disclosed.
Rebuttal Phase
After the initial review phase, authors will have an opportunity to respond to the reviews. Responses should be succinct and focus on factual and technical points. Authors are not required to submit a rebuttal; papers may be accepted and receive awards without one.
Awards
At most three awards will be given:
- An overall best paper award
- A best student paper award, for which the primary author must be a full-time student
- An elegant paper award, given to a submission that is exceptionally concise, well written, or otherwise elegant
Contact
Questions about submissions should be sent to the program chairs at alt2027pc@gmail.com.