An Alternative to Backpropagation

Add to your list(s) Download to your calendar using vCal

• Varun Jain
• Wednesday 14 May 2025, 16:00-16:30
• Cambridge University Engineering Department, CBL Seminar room BE4-38. For directions see http://learning.eng.cam.ac.uk/Public/Directions.

If you have a question about this talk, please contact Cat Spencer.

For decades, backpropagation has been the backbone of deep learning. However, it is biologically implausible, computationally demanding, and sometimes inefficient. This has spurred the search for alternative training paradigms. I will introduce a novel gradient‑free approach in which the objective is to align the distribution of a network’s linear regions with that of the training data. On CIFAR ‑10, this approach surpasses the performance achieved by Adam, which demonstrates that carefully structuring a model’s linear regimes can provide a powerful alternative to traditional gradient‑based methods.

This talk is part of the CBL Research Talks series.

This talk is included in these lists:

• Cambridge University Engineering Department, CBL Seminar room BE4-38. For directions see http://learning.eng.cam.ac.uk/Public/Directions

Note that ex-directory lists are not shown.