Zero-shot LLM-guided Counterfactual Generation for Text

arXiv:2405.04793v1 Announce Type: cross
Abstract: Counterfactual examples are frequently used for model development and evaluation in many natural language processing (NLP) tasks. Although methods for automated counterfactual generation have been explored, such methods depend on models such as pre-trained language models that are then fine-tuned on auxiliary, often task-specific datasets. Collecting and annotating such datasets for counterfactual generation is labor intensive and therefore, infeasible in practice. Therefore, in this work, we focus on a novel problem setting: \textit{zero-shot counterfactual generation}. …

abstract arxiv automated counterfactual cs.ai cs.cl cs.lg datasets development evaluation examples language language models language processing llm model development natural natural language natural language processing nlp processing tasks text type zero-shot