The first Workshop on
Scalable and Efficient Artificial Intelligence Systems (SEAS)
March 4, 2025 @ AAAI 2025 in Philadelphia, Pennsylvania, USA
Venue: Pennsylvania Convention Center




Accepted Papers

Paper Title Authors
ssProp: Energy-Efficient Training for Convolutional Neural Networks with Scheduled Sparse Back Propagation Lujia Zhong (University of Southern California); Shuo Huang (University of Southern California); Yonggang Shi (University of Southern California)
Knowledge Distillation with Training Wheels Guanlin Liu (University of Califonia, Davis); Anand Ramachandran (Amazon); Tanmay Gangwani (Amazon); Yan Fu (Amazon); Abhinav Sethy (Amazon)
PickLLM: Context-Aware RL-Assisted Large Language Model Routing Dimitrios Sikeridis (Broadcom); Dennis Ramdass (Broadcom); Pranay Pareek (Broadcom)
ZNorm: Z-Score Gradient Normalization Accelerating Skip-Connected Network Training without Architectural Modification Juyoung Yun (Stony Brook University)
The Impact of Model Scaling on Seen and Unseen Language Performance Rhitabrat Pokharel (Portland State University); Ameeta Agrawal (Portland State University)
Information Consistent Pruning: How to Efficiently Search for Sparse Networks? Soheil Gharatappeh (University of Maine); Salimeh Sekeh (San Diego State University)
Efficient LLM Inference with I/O-Aware Partial KV Cache Recomputation Chaoyi Jiang (University of Southern California); Lei Gao (University of Southern California); Hossein Entezari Zarch (University of Southern California); Murali Annavaram (University of Southern California)
Differentiated Aggregation to Improve Generalization in Federated Learning Peyman Gholami (University of Illinois at Chicago); Hulya Seferoglu (University of Illinois at Chicago)
Efficient Image Similarity Search with Quadtrees Yifan Zhang (Lehigh University)



Instructions of Publication Preparation

We plan to publish the accepted papers with the Springer CCIS series. For the authors of the accepted papers, please follow these instructions:

  1. Please format your paper using the official Springer template and follow the instructions on the Springer author guideline and author information.
  2. Please also submit all source files (compressed as a .zip file) together with the compiled .pdf file and the signed License to Publish Agreement via CMT.
  3. If the camera-ready paper, the corresponding LaTeX source file, and the license to publish agreement are not submitted in the correct format on time, then the publication of the paper cannot be guaranteed.
  4. Deadline: Submit all required files by Jan 17, 2025 (11:59 PM PST).

Contact Info

E-mail: seas25 AT googlegroups.com