True Few-Shot Learning with Language Models

TL;DR

This paper investigates the true few-shot learning capabilities of pretrained language models without relying on held-out examples for tuning, revealing that current evaluation methods overestimate their effectiveness.

Contribution

The study evaluates model selection criteria in true few-shot settings and shows they perform poorly compared to random selection, challenging prior assumptions about LM capabilities.

Findings

Model selection criteria often perform worse than random.

Prior evaluations overestimate LM few-shot abilities.

Selection methods frequently prefer suboptimal models.

Abstract

Pretrained language models (LMs) perform well on many tasks even when learning from a few examples, but prior work uses many held-out examples to tune various aspects of learning, such as hyperparameters, training objectives, and natural language templates ("prompts"). Here, we evaluate the few-shot ability of LMs when such held-out examples are unavailable, a setting we call true few-shot learning. We test two model selection criteria, cross-validation and minimum description length, for choosing LM prompts and hyperparameters in the true few-shot setting. On average, both marginally outperform random selection and greatly underperform selection based on held-out examples. Moreover, selection criteria often prefer models that perform significantly worse than randomly-selected ones. We find similar results even when taking into account our uncertainty in a model's true performance during selection, as well as when varying the amount of computation and number of examples used for selection. Overall, our findings suggest that prior work significantly overestimated the true few-shot ability of LMs given the difficulty of few-shot model selection.