Does CLIPs Generalization Performance Mainly Stem from High Train-Test Similarity? | Robust Machine Learning – Max Planck Institute for Intelligent Systems

Institute Homepage

Institute Homepage Sign In

Research Overview

Comparative Vision Science

Does CLIP's Generalization Performance Mainly Stem from High Train-Test Similarity?

In Search of Forgotten Domain Generalization

Partial Success in Closing the Gap between Human and Machine Vision

Foundations of Machine Learning

Cross-Entropy Is All You Need To Invert the Data Generating Process

Contrastive Learning Inverts the Data Generating Process

Interaction Asymmetry: A General Principle for Learning Composable Abstractions

Mechanistic Interpretability

Measuring Per-Unit Interpretability at Scale Without Humans

Exemplary Natural Images explain CNN Activations better than State-Of-The-Art Feature Visualizations

How Well do Feature Visualizations Support Causal Understanding of CNN Activations?

Robust Machine Learning

Does CLIP’s Generalization Performance Mainly Stem from High Train-Test Similarity?

We developed a pruned version of CLIP’s training dataset to analyze its out-of-distribution (OOD) generalization, revealing that high train-test similarity is insufficient to explain CLIP's OOD performance~[

].