��������

The submission deadline for the "Learn to Compress & Compress to Learn" Workshop @ ISIT 2025 is now extended to March 28, 2025. For more detailed information, please visit and .

This workshop aims to unite experts from machine learning, computer science, and information theory to delve into the dual themes of learning-based compression and using compression as a tool for learning tasks.��

The program will feature invited talks �ڰ��dz�:��

*�� (University of Pennsylvania)

* ��(University of Cambridge)

* (Granica)

* (Apple)

* (Texas A&M University)

* (Chan Zuckerberg Initiative)

We invite researchers from machine learning, compression and related fields to submit their latest work to the workshop.��We welcome submissions of recent work that has been presented, published, or is currently under review elsewhere, if the authors opt out of publishing their paper on �������� Xplore.

All accepted papers will be presented as posters during the poster session. Some papers will also be selected for spotlight presentations. Topics of interest include but are not limited to:

• "Learn to Compress” – Advancing Compression with Learning

  • Learning-Based Data Compression:��New techniques for compressing data (e.g., images, video, audio), model weights, and emerging modalities (e.g., 3D content and AR/VR applications).

  • Efficiency for Large-Scale Foundation Models:��Accelerating training and inference for large-scale foundation models, particularly in distributed and resource-constrained settings

  • Theoretical Foundations of Neural Compression:��Fundamental limits (e.g., rate-distortion bounds), distortion/perceptual/realism metrics, distributed compression, compression without quantization (e.g., channel simulation, relative entropy coding), and stochastic/probabilistic coding techniques.

• "Compress to Learn” – Leveraging Principles of Compression to Improve Learning

  • Compression as a Tool for Learning:��Leveraging principles of compression and source coding to understand and improve learning and generalization.

  • Compression as a Proxy for Learning:��Understanding the information-theoretic role of compression in tasks like unsupervised learning, representation learning, and semantic understanding.

  • Interplay of Algorithmic Information Theory and Source Coding:��Exploring connections between Algorithmic Information Theory concepts (e.g., Kolmogorov complexity, Solomonoff induction) and emerging source coding methods.

Important Dates

• Paper submission deadline: [Extended]��March 28, 2025��(11:59 PM AoE, Anywhere on Earth).
• Decision notification:��April 18, 2025
• Camera-ready paper deadline:��May 1, 2025
• Workshop date:��June 26, 2025
  ��

Organizing Committee

* ��(NYU)

* (University of Cambridge / Imperial College London)

* (Imperial College London)

* (NYU)��

��