Challenge: Pretrained language models have achieved super-human performances on many Machine Reading Comprehension (MRC) benchmarks.
Approach: They propose to fine-tune three state-of-the-art language models on SQuAD 1.1 or SQu AD 2.0 and then evaluate their robustness under adversarial attacks.
Outcome: The proposed model is able to perform better under adversarial attacks than model fine-tuned on SQuAD 1.1 or 2.0.

Similar Papers

Evaluating Neural Model Robustness for Machine Comprehension (2021.eacl-main)

Copied to clipboard

Challenge: evaluating model robustness to adversarial attacks can provide deeper understanding of how deep neural networks work and what kind of linguistic information is actually captured by neural networks.
Approach: They propose a method for strategic sentence-level perturbations to evaluate model robustness to adversarial attacks using character and word perturbations.
Outcome: The proposed model improves model performance during adversarial attacks by using ensembles and predicts errors in adversarials.
Know What You Don’t Know: Unanswerable Questions for SQuAD (P18-2)

Copied to clipboard

Challenge: Existing datasets focus on answerable questions or use automatically generated unanswerable questions that are easy to identify.
Approach: They propose a dataset that combines the Stanford Question Answering Dataset with 50,000 unanswerable questions written by crowdworkers to look similar to answerable ones.
Outcome: The proposed dataset looks similar to answerable questions on crowd-written questions . strong neural system that gets 86% F1 on SQuAD achieves only 66% F1.
Robust Machine Comprehension Models via Adversarial Training (N18-2)

Copied to clipboard

Challenge: Existing models for the Stanford Question Answering Dataset suffer from a 50% decrease in F1 score during adversarial evaluation based on AddSent.
Approach: They propose an alternative adversary-generation algorithm, AddSentDiverse, that significantly increases the variance within the adversarial training data by providing effective examples that punish the model for making certain superficial assumptions.
Outcome: The proposed algorithm can achieve a 36.5% increase in F1 score while maintaining performance on the regular SQuAD task.
SQuAD2-CR: Semi-supervised Annotation for Cause and Rationales for Unanswerability in SQuAD 2.0 (2020.lrec-1)

Copied to clipboard

Challenge: Existing models are brittle for adversarial perturbed questions, causing uncertainty . a dataset with annotations on unanswerable questions is not available to solve this problem .
Approach: They use crowdsourced annotations to annotate unanswerable questions . they also annotated which part of the question causes unanswered questions a .
Outcome: The proposed dataset can be used to improve model interpretation, authors say . they find that existing models are brittle for adversarial perturbed questions .
Benchmarking Robustness of Machine Reading Comprehension Models (2021.findings-acl)

Copied to clipboard

Challenge: Existing benchmarks only evaluate models' robustness under test-time perturbations or adversarial attacks.
Approach: They propose a model-agnostic benchmark to evaluate models' robustness under adversarial attacks.
Outcome: The proposed model-agnostic benchmark evaluates models under four different types of adversarial attacks.
On the Robustness of Reading Comprehension Models to Entity Renaming (2022.naacl-main)

Copied to clipboard

Challenge: SpanBERT model is more robust than RoBERTa, despite having similar accuracy on unperturbed test data.
Approach: They propose a pipeline to replace entity names with names from a variety of sources.
Outcome: The proposed model performs worse when entities are renamed, the authors show . SpanBERT, which is pretrained with span-level masking, is more robust than RoBERTa .
Towards Robust Extractive Question Answering Models: Rethinking the Training Methodology (2024.findings-emnlp)

Copied to clipboard

Challenge: Existing models lack robustness against distribution shifts and adversarial attacks when training on unanswerable questions in EQA datasets.
Approach: They propose a novel loss function for the EQA problem to improve the robustness of extractive question answering models by adding adversarial questions to a crowdsourcing process.
Outcome: The proposed method maintains in-domain performance while improving on out-of-domain datasets.
Using Adversarial Attacks to Reveal the Statistical Bias in Machine Reading Comprehension Models (2021.acl-short)

Copied to clipboard

Challenge: Pre-trained language models have achieved human-level performance on many Machine Reading Comprehension (MRC) tasks, but it remains unclear whether these models truly understand language or answer questions by exploiting statistical biases in datasets.
Approach: They propose a method to attack MRC models by exposing statistical biases in a RACE dataset and propose an augmented training method that can greatly reduce models’ statistical bias.
Outcome: The proposed method can reduce models’ statistical biases from human-level performance to chance-level.
Learning Invariant Representation Improves Robustness for MRC Models (2022.findings-emnlp)

Copied to clipboard

Challenge: Existing approaches to improve machine reading comprehension models are vulnerable and not robust to adversarial examples.
Approach: They propose to construct positive example pairs which have same answer by augmentation and then introduce stability and contrastive loss to improve invariance of representation.
Outcome: The proposed approach boosts the robustness of QA models across different tasks and attack sets significantly and consistently.
Do We Know What We Don’t Know? Studying Unanswerable Questions beyond SQuAD 2.0 (2021.findings-emnlp)

Copied to clipboard

Challenge: Recent work has attempted to improve extractive QA performance by enriching the dataset with unanswerable questions.
Approach: They build an out-of-domain corpus of competitive and non-competitive questions . they compare the results with the results of the Recognizing Textual Entailments task .
Outcome: The proposed model fails even in the case of simpler questions . the proposed model can be used to address more realistic situations in reading comprehension .

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations