# Adversarial Training for Commonsense Inference

arXiv: https://arxiv.org/abs/2005.08156 · HTML: https://mlsanity.com/paper/2005.08156/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 May 2020.

## Built on by (most cited first, 13 of 13)

- [DAGN: Discourse-Aware Graph Network for Logical Reasoning](https://mlsanity.com/paper/2103.14349/) (2021)
- [An Overview Of Temporal Commonsense Reasoning and Acquisition](https://mlsanity.com/paper/2308.00002/) (2023)
- [LAMPAT: Low-Rank Adaption for Multilingual Paraphrasing Using Adversarial Training](https://mlsanity.com/paper/2401.04348/) (2024)
- [Targeted Adversarial Training for Natural Language Understanding](https://mlsanity.com/paper/2104.05847/) (2021)
- [How Does Adversarial Fine-Tuning Benefit BERT?](https://mlsanity.com/paper/2108.13602/) (2021)
- [Discourse-Aware Graph Networks for Textual Logical Reasoning](https://mlsanity.com/paper/2207.01450/) (2022)
- [Temporal Validity Change Prediction](https://mlsanity.com/paper/2401.00779/) (2024)
- [Neuro-Symbolic Artificial Intelligence: A Task-Directed Survey in the Black-Box Models Era](https://mlsanity.com/paper/2603.03177/) (2026)
- [Adversarial Training for Machine Reading Comprehension with Virtual Embeddings](https://mlsanity.com/paper/2106.04437/) (2021)
- [Multi-grained Evidence Inference for Multi-choice Reading Comprehension](https://mlsanity.com/paper/2310.18070/) (2023)
- [OCHADAI-KYOTO at SemEval-2021 Task 1: Enhancing Model Generalization and Robustness for Lexical Complexity Prediction](https://mlsanity.com/paper/2105.05535/) (2021)
- [elBERto: Self-supervised Commonsense Learning for Question Answering](https://mlsanity.com/paper/2203.09424/) (2022)
- [OCHADAI at SemEval-2022 Task 2: Adversarial Training for Multilingual Idiomaticity Detection](https://mlsanity.com/paper/2206.03025/) (2022)

## Abstract (arXiv)

We propose an AdversariaL training algorithm for commonsense InferenCE (ALICE). We apply small perturbations to word embeddings and minimize the resultant adversarial risk to regularize the model. We exploit a novel combination of two different approaches to estimate these perturbations: 1) using the true label and 2) using the model prediction. Without relying on any human-crafted features, knowledge bases, or additional datasets other than the target datasets, our model boosts the fine-tuning performance of RoBERTa, achieving competitive results on multiple reading comprehension datasets that require commonsense inference.