Test-Time Training on Nearest Neighbors for Large Language Models

TL;DR

This paper introduces a test-time training method for large language models that retrieves nearest neighbors and fine-tunes the model on them, significantly improving performance without increasing input length or computational cost.

Contribution

It presents a novel test-time training approach using nearest neighbors, demonstrating substantial performance gains across multiple language modeling tasks.

Findings

Retrieving and training on 20 neighbors improves model performance.

Test-time training narrows the gap between small and large language models.

High-quality, large index is crucial for effectiveness.

Abstract

Many recent efforts augment language models with retrieval, by adding retrieved data to the input context. For this approach to succeed, the retrieved data must be added at both training and test time. Moreover, as input length grows linearly with the size of retrieved data, cost in computation and memory grows quadratically for modern Transformers. To avoid these complications, we simply fine-tune the model on retrieved data at test time, using its standard training setup. We build a large-scale distributed index based on text embeddings of the Pile dataset. For each test input, our system retrieves its neighbors and fine-tunes the model on their text. Surprisingly, retrieving and training on as few as 20 neighbors, each for only one gradient iteration, drastically improves performance across more than 20 language modeling tasks in the Pile. For example, test-time training with nearest neighbors significantly narrows the performance gap between a small GPT-2 and a GPT-Neo model more than 10 times larger. Sufficient index quality and size, however, are necessary. Our work establishes a first baseline of test-time training for language modeling.