Tomasz Trzciński Co-Authors Award-Winning Paper at NeurIPS 2025

We’re excited to share that Tomasz Trzciński, ELLIS Fellow and director of the ELLIS Unit Warsaw, is a co-author of a paper that received a Best Paper Award at NeurIPS 2025.

NeurIPS is one of the most prestigious global conferences in machine learning and AI, and this year’s Best Paper Awards highlight breakthrough contributions selected from both the Main Track and the Datasets & Benchmark Track.

About the awarded paper
The recognized work, titled “1000 Layer Networks for Self-Supervised RL: Scaling Depth Can Enable New Goal-Reaching Capabilities” (authors: Kevin Wang , Ishaan Javali, Michał Bortkiewicz, Tomasz Trzcinski, Benjamin Eysenbach), explores how extremely deep architectures can unlock new possibilities in self-supervised reinforcement learning. The results demonstrate how scaling depth transforms the goal-reaching abilities of RL agents, adding an important piece to the broader conversation about scaling laws in modern AI.

Why this matters
This award is an excellent recognition of the scientific strength emerging from the Warsaw ELLIS community and a great example of impactful, internationally visible research with ELLIS involvement. Huge congratulations to Tomek and the whole author team!