Skip to yearly menu bar Skip to main content

Spotlight Poster

Intriguing Properties of Generative Classifiers

Priyank Jaini · Kevin Clark · Robert Geirhos

Halle B #83
[ ]
Thu 9 May 7:30 a.m. PDT — 9:30 a.m. PDT


What is the best paradigm to recognize objects---discriminative inference (fast but potentially prone to shortcut learning) or using a generative model (slow but potentially more robust)? We build on recent advances in generative modeling that turn text-to-image models into classifiers. This allows us to study their behavior and to compare them against discriminative models and human psychophysical data.We report four intriguing emergent properties of generative classifiers: they show a record-breaking human-like shape bias (99% for Imagen), near human-level out-of-distribution accuracy, state-of-the-art alignment with human classification errors, and they understand certain perceptual illusions. Our results indicate that while the current dominant paradigm for modeling human object recognition is discriminative inference, zero-shot generative models approximate human object recognition data surprisingly well.

Chat is not available.