We develop a methodology for analyzing language model task performance at the
individual example level based on training data density estimation. Experiments
with paraphrasing as a controlled intervention on finetuning data demonstrate
that increasing the support in the training distribution for specific test
queries results in a measurable increase in density, which is also a
significant predictor of the performance increase caused by the intervention.
Experiments with pretraining data demonstrate that we can explain a significant
fraction of the variance in model perplexity via density measurements. We
conclude that our framework can provide statistical evidence of the dependence
of a target model's predictions on subsets of its training data, and can more
generally be used to characterize the support (or lack thereof) in the training
data for a given test task.

我们开发了一种基于训练数据密度估计的分析语言模型任务性能的方法。通过对有控制介入的细调数据进行释义实验，我们发现支持特定测试查询的训练分布的增加可导致密度的可测增加，这也是干预引起的性能增加的重要预测因素。通过预训练数据的实验，我们可以通过密度测量解释模型困惑度的相当大的方差。我们得出结论，我们的框架能够提供目标模型预测与其训练数据子集的依赖性的统计证据，并且可以更一般地用于表征给定测试任务的训练数据的支持或缺乏支持。