Call for Papers and Award

Important Dates

• Paper Submission Deadline: Monday, Mar 3, 2025 (23:59 AoE)

• Notification of Acceptance: Friday, March 28, 2025

• Camera-Ready Submission: Friday, May 16, 2025

• Workshop Date: June 13, 2025

All deadlines are at 23:59 AoE (Anywhere on Earth) time.

Submission instructions

Submission follows the ISC format: 

• https://cmt3.research.microsoft.com/CADL2025 

Conference Proceedings

The post-conference Workshop Proceedings will be published with Springer’s Lecture Notes in Computer Science (LNCS).
 

Upload your paper in PDF format, up to 12 pages (optional: two extra pages to incorporate reviewers comments). Use the LNCS formatting for your paper. More instructions can be found here: https://www.springer.com/gp/computer-science/lncs/conference-proceedings-guidelines?

List of Topics

• Optimization strategies for reducing energy consumption in deep learning

• Novel architectures and operators for data-intensive scenarios

• Distributed and efficient reinforcement learning algorithms

• Large-scale pre-training techniques for real-world applications

• Distributed training approaches and architectures

• Utilization of HPC and massively parallel architectures for deep learning

• Frameworks and optimization algorithms for training deep networks

• Model pruning, gradient compression, and quantization techniques for reducing computational complexity

• Methods to reduce memory and data transmission footprints

• Differentiable metrics for estimating computational costs, energy consumption, and power usage of models

• Design, implementation, and application of hardware accelerators for deep learning

• Efficient and cost-effective models and methods promoting diversity and inclusivity in deep learning

• Acceleration of training and inference for large language models (e.g., GPT) and other generative models

• Efficient deep learning for edge and embedded devices

• Green AI: Sustainable practices and carbon footprint reduction in AI research and deployment

• Scalable and efficient multi-modal learning architectures

• Optimization techniques for transformer-based models in various domains

• Efficient federated learning algorithms and architectures

• Computational aspects of AI safety and robustness

• Resource-efficient few-shot and zero-shot learning techniques

• This revised list maintains the original focus while adding a few relevant topics that reflect recent advancements and concerns in the field. The additions (items 14-20) broaden the scope to include emerging areas of interest in computational aspects of deep learning.

Acknowledgment 

The Microsoft CMT service will be used for managing the peer-reviewing process for this workshop. This service was provided for free by Microsoft and they bore all expenses, including costs for Azure cloud services as well as for software development and support.

Award

We are pleased to announce that the winner of the Best Paper award is ‘Direct Feedback Alignment for Recurrent Neural Networks’.

Congratulations to Andrea Cossu and the other authors for the outstanding contribution and impact of their research!