CNT (Conditioning on Noisy Targets): A new Algorithm for Leveraging Top-Down Feedback
This addresses the challenge of efficient learning in supervised tasks like imitation learning and classification, though it appears incremental as it builds on existing regularization techniques.
The paper tackles the problem of improving supervised learning by proposing a novel regularizer called Conditioning on Noisy Targets (CNT), which conditions models on noisy versions of targets at varying noise levels to provide top-down feedback, resulting in benefits such as focusing on simpler sub-problems and progressive learning from easy to hard examples.
We propose a novel regularizer for supervised learning called Conditioning on Noisy Targets (CNT). This approach consists in conditioning the model on a noisy version of the target(s) (e.g., actions in imitation learning or labels in classification) at a random noise level (from small to large noise). At inference time, since we do not know the target, we run the network with only noise in place of the noisy target. CNT provides hints through the noisy label (with less noise, we can more easily infer the true target). This give two main benefits: 1) the top-down feedback allows the model to focus on simpler and more digestible sub-problems and 2) rather than learning to solve the task from scratch, the model will first learn to master easy examples (with less noise), while slowly progressing toward harder examples (with more noise).