Simple Contrastive Representation Adversarial Learning for NLP Tasks

TL;DR

This paper introduces a novel adversarial contrastive learning framework for NLP that enhances model robustness and generalization by generating challenging examples in embedding space, outperforming existing methods on benchmarks.

Contribution

It proposes two frameworks, SCAL and USCAL, combining adversarial training with contrastive learning for NLP, improving robustness and performance on various tasks.

Findings

Supervised SCAL outperforms BERT_base by 1.75% on GLUE.

Unsupervised USCAL achieves 77.29% on STS tasks.

State-of-the-art robustness on multiple adversarial NLI datasets.

Abstract

Self-supervised learning approach like contrastive learning is attached great attention in natural language processing. It uses pairs of training data augmentations to build a classification task for an encoder with well representation ability. However, the construction of learning pairs over contrastive learning is much harder in NLP tasks. Previous works generate word-level changes to form pairs, but small transforms may cause notable changes on the meaning of sentences as the discrete and sparse nature of natural language. In this paper, adversarial training is performed to generate challenging and harder learning adversarial examples over the embedding space of NLP as learning pairs. Using contrastive learning improves the generalization ability of adversarial training because contrastive loss can uniform the sample distribution. And at the same time, adversarial training also enhances the robustness of contrastive learning. Two novel frameworks, supervised contrastive adversarial learning (SCAL) and unsupervised SCAL (USCAL), are proposed, which yields learning pairs by utilizing the adversarial training for contrastive learning. The label-based loss of supervised tasks is exploited to generate adversarial examples while unsupervised tasks bring contrastive loss. To validate the effectiveness of the proposed framework, we employ it to Transformer-based models for natural language understanding, sentence semantic textual similarity and adversarial learning tasks. Experimental results on GLUE benchmark tasks show that our fine-tuned supervised method outperforms BERT$_{base}$ over 1.75\%. We also evaluate our unsupervised method on semantic textual similarity (STS) tasks, and our method gets 77.29\% with BERT$_{base}$. The robustness of our approach conducts state-of-the-art results under multiple adversarial datasets on NLI tasks.