We present an ngram model-based logit scaling technique that effectively transfers extreme subword stylistic variation to large language models at inference time. We demonstrate its efficacy by tracking the perplexity of generated text with respect to the ngram interpolated and original versions of an evaluation model. Minimizing the former measure while the latter approaches the perplexity of a text produced by a target author or character lets us select a sufficient degree of adaptation while retaining fluency.

翻译：本文提出一种基于N-gram模型的对数概率缩放技术，该技术能够在推理阶段有效地将极端子词风格变化迁移至大型语言模型。我们通过对比生成文本在N-gram插值版本与原始版本评估模型下的困惑度来验证其有效性。在保持目标作者或角色文本原始困惑度的前提下，通过最小化插值模型的困惑度指标，使我们能够在保持文本流畅性的同时选择充分的适应程度。