论文标题
核心：检索反事实数据的框架
CORE: A Retrieve-then-Edit Framework for Counterfactual Data Generation
论文作者
论文摘要
反事实数据增强（CDA） - 即，在培训过程中添加最低扰动的输入 - 有助于减少模型依赖虚假相关性并改善对分布（OOD）数据的概括。关于产生反事实的先前工作仅考虑限制扰动类别，从而限制了其有效性。我们通过检索和编辑（核心）介绍了反事实的生成，这是一个检索杰出的生成框架，用于为CDA创造各种反事实扰动。对于每个培训示例，Core首先使用学习的双重编码器对与任务相关的无标记的文本语料库进行密集检索，并提取相关的反事实摘录。然后，Core将这些提示纳入具有很少的学习能力的大型语言模型，以进行反事实编辑。调节语言模型编辑自然发生的数据会导致各种扰动。自然语言推理和情感分析的实验基准表明，与其他DA方法相比，核心反事实在改善对OOD数据的概括方面更有效。我们还表明，可以使用核心检索框架来鼓励手动撰写的扰动多样性
Counterfactual data augmentation (CDA) -- i.e., adding minimally perturbed inputs during training -- helps reduce model reliance on spurious correlations and improves generalization to out-of-distribution (OOD) data. Prior work on generating counterfactuals only considered restricted classes of perturbations, limiting their effectiveness. We present COunterfactual Generation via Retrieval and Editing (CORE), a retrieval-augmented generation framework for creating diverse counterfactual perturbations for CDA. For each training example, CORE first performs a dense retrieval over a task-related unlabeled text corpus using a learned bi-encoder and extracts relevant counterfactual excerpts. CORE then incorporates these into prompts to a large language model with few-shot learning capabilities, for counterfactual editing. Conditioning language model edits on naturally occurring data results in diverse perturbations. Experiments on natural language inference and sentiment analysis benchmarks show that CORE counterfactuals are more effective at improving generalization to OOD data compared to other DA approaches. We also show that the CORE retrieval framework can be used to encourage diversity in manually authored perturbations
