Therefore, long-tailed classification is the key to deep learning at scale. However, existing methods are mainly based on reweighting/re-sampling heuristics that lack a fundamental theory. In this paper, weestablish acausal inference framework,which notonlyunravelsthewhysof previous methods, but also derives a new principled solution.

Deceptive agents are a challenge for the safety, trustworthiness, and cooperation of AI systems.

The unified perspective on memory mechanisms allows us to derive these closed-form solutions with thesamestrategy,providedinAppendices D.1,D.2.

Shift invariance is a critical property of CNNs that improves performance on classification.

The ubiquitous and demonstrably suboptimal choice of resizing images to a fixed resolution before processing them with computer vision models has not yet been successfully challenged.