Iterative Counterfactual Data Augmentation
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910844112076800 |
|---|---|
| author | Plyler, Mitchell Chi, Min |
| author_facet | Plyler, Mitchell Chi, Min |
| contents | Counterfactual data augmentation (CDA) is a method for controlling information or biases in training datasets by generating a complementary dataset with typically opposing biases. Prior work often either relies on hand-crafted rules or algorithmic CDA methods which can leave unwanted information in the augmented dataset. In this work, we show iterative CDA (ICDA) with initial, high-noise interventions can converge to a state with significantly lower noise. Our ICDA procedure produces a dataset where one target signal in the training dataset maintains high mutual information with a corresponding label and the information of spurious signals are reduced. We show training on the augmented datasets produces rationales on documents that better align with human annotation. Our experiments include six human produced datasets and two large-language model generated datasets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_18249 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Iterative Counterfactual Data Augmentation Plyler, Mitchell Chi, Min Machine Learning Computation and Language Information Theory Counterfactual data augmentation (CDA) is a method for controlling information or biases in training datasets by generating a complementary dataset with typically opposing biases. Prior work often either relies on hand-crafted rules or algorithmic CDA methods which can leave unwanted information in the augmented dataset. In this work, we show iterative CDA (ICDA) with initial, high-noise interventions can converge to a state with significantly lower noise. Our ICDA procedure produces a dataset where one target signal in the training dataset maintains high mutual information with a corresponding label and the information of spurious signals are reduced. We show training on the augmented datasets produces rationales on documents that better align with human annotation. Our experiments include six human produced datasets and two large-language model generated datasets. |
| title | Iterative Counterfactual Data Augmentation |
| topic | Machine Learning Computation and Language Information Theory |
| url | https://arxiv.org/abs/2502.18249 |