There is Strength in Numbers: Avoiding the Hypothesis-Only Bias in Natural Language Inference via Ensemble Adversarial Training

By Joe Stacey, Pasquale Minervini, Haim Dubossarsky, Sebastian Riedel, Tim Rocktaschel. 2020

54 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Natural Language Processing

Citations in the corpus

Are We Done with MMLU? (2024, neutral): “In NLP, similar issues have been found in summarisation [20, 21] and natural language inference [22, 23, 24, 25] benchmarks.”
Supervising Model Attention with Human Explanations for Robust Natural Language Inference (2021, methodology): “Other strategies to prevent models learning from artefacts include using adversarial training with gradient reversal to mitigate the hypothesis-only bias (Belinkov et al. 2019a,b; Stacey et al. 2020), using data-augmentation (Min et al.”
Generating Data to Mitigate Spurious Correlations in Natural Language Inference Datasets (2022, methodology): “Furthermore, we compare our method to strong debiasing strategies from the literature (Be-linkov et al., 2019b; Stacey et al., 2020; Clark et al., 2019; Karimi Mahabadi et al., 2020; Utama et al., 2020; Sanh et al., 2021; Ghaddar et al., 2021).”
Learning from others' mistakes: Avoiding dataset biases without modeling them (2020, background): “Several active areas of research tackle this challenge by adversarial training (Belinkov et al., 2019; Stacey et al., 2020), example forgetting (Yaghoobzadeh et al., 2019) and dynamic loss adjustment (Cad`ene et al., 2019).”
Natural Language Inference with a Human Touch: Using Human Explanations to Guide Model Attention (2021, methodology): “…models learning from artefacts include adversarial training using gradient reversal to mitigate the hypothesis-only bias (Belinkov et al., 2019a; Stacey et al., 2020), using data-augmentation (Min et al., 2020; Minervini and Riedel, 2018), ﬁne-tuning on minority examples (Yaghoobzadeh et al.,…”