Seeing Is Believing: Brain-Inspired Modular Training for Mechanistic Interpretability
We introduce Brain-Inspired Modular Training (BIMT), a method for making neural networks more modular and interpretable. Inspired by brains, BIMT embeds neurons in a geometric space and augments the loss function with a cost proportional to the length of each neuron connection. This is inspired by t...
Main Authors: | Ziming Liu, Eric Gan, Max Tegmark |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-12-01
|
Series: | Entropy |
Subjects: | |
Online Access: | https://www.mdpi.com/1099-4300/26/1/41 |
Similar Items
-
Perspective and Development of Brain-Inspired Swarm Navigation
by: Li Weibin, Yang Donghao, Li Zhanbin, Hou Biao, Jiao Licheng
Published: (2023-10-01) -
Editorial: Brain-inspired navigation and sensing for robots
by: Hui Zhao, et al.
Published: (2023-11-01) -
Effects of antioxidants on diabetic kidney diseases: mechanistic interpretations and clinical assessment
by: Yuting Sun, et al.
Published: (2023-01-01) -
Barriers to the Adoption of Modular Construction in Portugal: An Interpretive Structural Modeling Approach
by: Adriana Machado Ribeiro, et al.
Published: (2022-09-01) -
Editorial: Novel materials, devices and solutions for brain-inspired sensing and computing
by: Christian Monzio Compagnoni, et al.
Published: (2022-11-01)