Adaptation Approaches for Nearest Neighbor Language Models

TL;DR

This paper explores various adaptation strategies for $k$NN-LMs, including model, datastore, and score adaptations, demonstrating significant perplexity improvements across multiple domains.

Contribution

It introduces and evaluates combined adaptation methods for $k$NN-LMs, filling a gap in domain adaptation research for semi-parametric language models.

Findings

Combined adaptation outperforms baselines in perplexity reduction.

Perplexity improved by 17.1% and 16% over baseline methods.

Evaluation conducted across seven diverse domains.

Abstract

Semi-parametric Nearest Neighbor Language Models ($k$NN-LMs) have produced impressive gains over purely parametric LMs, by leveraging large-scale neighborhood retrieval over external memory datastores. However, there has been little investigation into adapting such models for new domains. This work attempts to fill that gap and suggests the following approaches for adapting $k$NN-LMs -- 1) adapting the underlying LM (using Adapters), 2) expanding neighborhood retrieval over an additional adaptation datastore, and 3) adapting the weights (scores) of retrieved neighbors using a learned Rescorer module. We study each adaptation strategy separately, as well as the combined performance improvement through ablation experiments and an extensive set of evaluations run over seven adaptation domains. Our combined adaptation approach consistently outperforms purely parametric adaptation and zero-shot ($k$NN-LM) baselines that construct datastores from the adaptation data. On average, we see perplexity improvements of 17.1% and 16% for these respective baselines, across domains.