通过隐式自一致性提升大模型生成事实准确性
Integrative Decoding: Improve Factuality via Implicit Self-consistency
- 用多轮采样结果拼接输入，动态聚合预测选择下一词
- 在TruthfulQA等数据集上提升11.2%以上事实准确率
- 适合开放生成任务，无需特定格式限制
基于自一致性的方法通过多次采样生成多个输出，并选择最一致的结果作为最终响应，显著提升了大语言模型的事实准确性。然而，现有方法对任务格式要求严格，适用范围受限。本文提出整合解码（Integrative Decoding, ID），使自一致性适用于开放式生成任务。ID通过将先前采样的响应拼接到输入前，同时处理多组输入，在每一步解码中聚合所有对应预测来选择下一个词。该方法在解码目标中隐式引入自一致性。大量实验表明，ID在多种语言模型上均显著提升事实性，在TruthfulQA（+11.2%）、Biographies（+15.4%）和LongFact（+8.5%）基准上表现优异。性能随采样数量增加而持续提升，显示其可通过重复采样扩展潜力。
原文摘要 · Abstract (English)
Self-consistency-based approaches, which involve repeatedly sampling multiple outputs and selecting the most consistent one as the final response, prove to be remarkably effective in improving the factual accuracy of large language models. Nonetheless, existing methods usually have strict constraints on the task format, largely limiting their applicability. In this paper, we present Integrative Decoding (ID), to unlock the potential of self-consistency in open-ended generation tasks. ID operates by constructing a set of inputs, each prepended with a previously sampled response, and then processes them concurrently, with the next token being selected by aggregating of all their corresponding predictions at each decoding step. In essence, this simple approach implicitly incorporates self-consistency in the decoding objective. Extensive evaluation shows that ID consistently enhances factuality over a wide range of language models, with substantial improvements on the TruthfulQA (+11.2%), Biographies (+15.4%) and LongFact (+8.5%) benchmarks. The performance gains amplify progressively as the number of sampled responses increases, indicating the potential of ID to scale up with repeated sampling.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。