paper-with-me

Papers

An Investigation of the (In)effectiveness of Counterfactually Augmented Data

2021-07-01 · ACL 2022 5 · Nitish Joshi, He He

While pretrained language models achieve excellent performance on natural language understanding benchmarks, they tend to rely on spurious correlations and generalize poorly to out-of-distribution (OOD) data. Recent work has explored using counterfactually-augmented data (CAD) -- data generated by minimally perturbing examples to flip the ground-truth label -- to identify robust features that are invariant under distribution shift. However, empirical results using CAD for OOD generalization have been mixed. To explain this discrepancy, we draw insights from a linear Gaussian model and demonstrate the pitfalls of CAD. Specifically, we show that (a) while CAD is effective at identifying robust features, it may prevent the model from learning unperturbed robust features; and (b) CAD may exacerbate existing spurious correlations in the data. On two crowdsourced CAD datasets, our results show that the lack of perturbation diversity limits their effectiveness on OOD generalization, calling for innovative crowdsourcing procedures to elicit diverse perturbation of examples.

📄 PDF Abstract BibTeX arXiv:2107.00753

Code (1)

joshinh/investigation-cad 공식 구현

Tasks

DiversityNatural Language Understanding

Methods 이 논문이 사용한 방법론

FLIP https://developer.nvidia.com/blog/flip-a-difference-evaluator-for-alternating-images/

Similar Papers 제목 키워드 기반

An Investigation of the (In)effectiveness of Counterfactually Augmented Data

2021-08-17 · ACL ARR November 2021 11 · Anonymous

While pretrained language models achieve excellent performance on natural language understanding benchmarks, they tend to rely on spurious correlations and generalize poorly to out-of-distribution (OOD) data. Recent work…

DiversityNatural Language Understanding

Counterfactually-Augmented SNLI Training Data Does Not Yield Better Generalization Than Unaugmented Data

2020-10-09 · EMNLP (insights) 2020 11 · William Huang, Haokun Liu, Samuel R. Bowman

A growing body of work shows that models exploit annotation artifacts to achieve state-of-the-art performance on standard crowdsourced benchmarks---datasets collected from crowdworkers to create an evaluation task---whil…

counterfactualNatural Language InferenceNatural Language UnderstandingReading Comprehension

Rethinking Video-Text Understanding: Retrieval from Counterfactually Augmented Data

2024-07-18 · Wufei Ma, Kai Li, Zhongshi Jiang, Moustafa Meshry 외

Recent video-text foundation models have demonstrated strong performance on a wide variety of downstream video understanding tasks. Can these video-text models genuinely understand the contents of natural videos? Standar…

Language ModellingLarge Language ModelRetrievalVideo Understanding

Counterfactually Augmented Data and Unintended Bias: The Case of Sexism and Hate Speech Detection

2022-05-09 · NAACL 2022 7 · Indira Sen, Mattia Samory, Claudia Wagner, Isabelle Augenstein

Counterfactually Augmented Data (CAD) aims to improve out-of-domain generalizability, an indicator of model robustness. The improvement is credited with promoting core features of the construct over spurious artifacts th…

Hate Speech Detection

PairCFR: Enhancing Model Training on Paired Counterfactually Augmented Data through Contrastive Learning

2024-06-09 · Xiaoqi Qiu, Yongjie Wang, Xu Guo, Zhiwei Zeng 외

Counterfactually Augmented Data (CAD) involves creating new data samples by applying minimal yet sufficient modifications to flip the label of existing data samples to other classes. Training with CAD enhances model robu…

Contrastive Learningcounterfactual