MLSS^R&S 2026

Overview

MLSS^R&S 2026 is a five day summer school dedicated to modern topics in machine learning with a strong emphasis on reliability, safety, and robustness of ML systems. The event is organized by the ML in PL Association and takes place in Krakow, Poland, from 29 June to 3 July 2026. Attendees will engage with invited lectures delivered by researchers and industry practitioners who are actively contributing to the fields of AI safety, reliability, and trustworthy AI. The setting is Krakow, a historic city known for its scholarly traditions, offering a platform to learn from world-renowned experts while networking with peers.

Agenda

The program centers on a lecture-based format with invited speakers from academia and industry. Each day features multiple lecture blocks, a dedicated poster session, and ample time for questions and discussion. The schedule typically includes breaks for coffee and lunch to foster interaction among participants and speakers. The previous edition showcased daily structures highlighted by a mix of talks and poster activities, emphasizing opportunities for attendees to present their work and receive feedback. While the 2026 agenda has not been fully published yet, the outline mirrors the established format: invited lectures, poster sessions, and structured daily programming designed to promote deep understanding of ML safety, robustness, and reliability.

Speakers

MLSS^R&S 2026 will feature a lineup of leading researchers and practitioners focusing on reliability and safety in machine learning. The page mentions a rotating roster of invited speakers, with initial profiles including:

• Franziska Boenisch (CISPA Helmholtz Center)
• Dominik Janzing (Amazon)
• Wojciech Samek (TU Berlin / Fraunhofer HHI)
• Adam Dziedzic (CISPA Helmholtz Center)
• Randall Balestriero (Brown University / Meta AI Research)
• Fazl Barez (University of Oxford)
• Alexandra Gomez-Villa (Universitat Autònoma de Barcelona)
• Anna Sztyber-Betley (Warsaw University of Technology / Truthful AI)
• Jan Betley (Truthful AI)
• Tomasz Michalak (IDEAS RI / Ellis Unit Warsaw)
• Wojciech Kusa (NASK)
• Kamil Mamak (Jagiellonian University)
• Adel Bibi (University of Oxford)
• Bartosz Zieliński (Jagiellonian University)
• Christian Schroeder de Witt (University of Oxford)

The organizers note that the full speaker list will be announced progressively. The descriptions emphasize expertise in explainability, causal inference, safety evaluations, robust ML, governance, and ethical considerations in AI.

Venue and Format

The event is an offline gathering hosted by Jagiellonian University in Krakow. On the first day, activities are planned at Collegium Novum, with subsequent days spanning the Faculty of Mathematics and Computer Science. The venue information includes a city address and a map, indicating a well-organized on-site experience with room for lectures, discussion, and social interaction. Applications and participation details are conveyed through the registration page, and there are several policy documents available such as Terms of Participation, Privacy Policy, and Anti-Harassment Policy. The registration process supports early bird and regular applications with deadlines, and attendees are advised to have a background in machine learning fundamentals. The event also provides contact information for inquiries and sponsor engagement opportunities for organizations wishing to support ML talent development.

Who should attend and FAQs

The intended audience includes PhD students, research-oriented Master students, and early-career researchers from academia and industry who are interested in reliability and safety of ML systems. Applicants are expected to have prior knowledge of ML fundamentals, including supervised and unsupervised learning, optimization, and probability, with a familiarity of mathematical tools commonly used in ML recommended. The site encourages attendees to engage with speakers, attend lectures, participate in poster sessions, and take part in networking events. For questions, attendees are directed to the provided contact email mlss@mlinpl.org and related social channels for updates and further information.

Summary

MLSS^R&S 2026 offers a comprehensive, on-site learning experience focused on the reliability and safety of machine learning, featuring invited lectures, interactive sessions, a poster track, and structured daily programming. It connects aspiring researchers with established experts while leveraging Krakow’s scholarly ambiance to foster collaboration and insight across academia and industry.