Timezone: »

Understanding prompt engineering does not require rethinking generalization
Victor Akinwande · Yiding Jiang · Dylan Sam · Zico Kolter
Event URL: https://openreview.net/forum?id=mwNnElcBvt »

Zero-shot learning in prompted visual-language models, the practice of crafting prompts to build classifiers without an explicit training process, shows an impressive performance in many settings. There also emerges a seemingly surprising fact: this method suffers relatively little from overfitting; i.e., when a prompt is manually engineered to achieve low error on a given training set (thus rendering the method no longer zero-shot), the approach still performs relatively well on held-out test data. In this paper, we show that we can explain such performance remarkably well via recourse to classical PAC-Bayes bounds. Specifically, we show that the discrete nature of prompts, combined with a PAC-Bayes prior given by a language model, results in generalization bounds that are \emph{remarkably} tight by the standards of the literature: for instance, the generalization bound of an ImageNet classifier is often within a few percentage points of the true test error. Indeed, we show that we can therefore \emph{greedily} search over the prompt space in such a framework, improving upon training performance while retaining the same bound. Furthermore, the bound is remarkably suitable for model selection: the models with the best bound typically also have the best test performance. This work thus provides a substantial justification for the widespread use of ``prompt engineering,'' even if it seems as though such methods could overfit a training set.

Author Information

Victor Akinwande (Carnegie Mellon University)
Yiding Jiang (Carnegie Mellon University)
Dylan Sam (Carnegie Mellon University)

ML PhD Student at Carnegie Mellon University

Zico Kolter (Carnegie Mellon University / Bosch Center for AI)

More from the Same Authors