Correcting Three Existing Beliefs on Mutual Information in Contrastive Learning

Kyungeun Lee, Jaeill Kim, Suhyun Kang, Wonjong Rhee

Abstract: Contrastive learning has played a pivotal role in the recent success of unsupervised representation learning. It has been commonly explained with instance discrimination and a mutual information loss, and some of the fundamental explanations are based on mutual information analysis. In this work, we develop new methods that enable rigorous analysis of mutual information in contrastive learning. Using the methods, we investigate three existing beliefs and show that they are incorrect. Based on the investigation results, we address two issues in the discussion section. In particular, we question if contrastive learning is indeed an unsupervised representation learning method because the current framework of contrastive learning relies on validation performance for tuning the augmentation design.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Unsupervised and Self-supervised learning