SWAG: A Large-Scale Adversarial Dataset for Grounded Commonsense Inference

  title={SWAG: A Large-Scale Adversarial Dataset for Grounded Commonsense Inference},
  author={Rowan Zellers and Yonatan Bisk and Roy Schwartz and Yejin Choi},
  • Rowan Zellers, Yonatan Bisk, +1 author Yejin Choi
  • Published in EMNLP 2018
  • Computer Science
  • Given a partial description like "she opened the hood of the car," humans can reason about the situation and anticipate what might come next ("then, she examined the engine. [...] Key Method To address the recurring challenges of the annotation artifacts and human biases found in many existing datasets, we propose Adversarial Filtering (AF), a novel procedure that constructs a de-biased dataset by iteratively training an ensemble of stylistic classifiers, and using them to filter the data.Expand Abstract
    262 Citations
    HellaSwag: Can a Machine Really Finish Your Sentence?
    • 97
    • PDF
    CODAH: An Adversarially Authored Question-Answer Dataset for Common Sense
    • 17
    • Highly Influenced
    • PDF
    G-DAUG: Generative Data Augmentation for Commonsense Reasoning
    • 10
    Generative Data Augmentation for Commonsense Reasoning
    • 2
    • PDF
    WINOGRANDE: An Adversarial Winograd Schema Challenge at Scale
    • 86
    • PDF
    C L ] 6 J un 2 01 9 CODAH : An Adversarially-Authored Question Answering Dataset for Common Sense
    • Highly Influenced
    CommonGen: A Constrained Text Generation Challenge for Generative Commonsense Reasoning
    • 13
    • PDF
    XCOPA: A Multilingual Dataset for Causal Commonsense Reasoning
    • 10
    • PDF


    Annotation Artifacts in Natural Language Inference Data
    • 367
    • PDF
    Mitigating Unwanted Biases with Adversarial Learning
    • 274
    • PDF
    Tackling the Story Ending Biases in The Story Cloze Test
    • 23
    • PDF
    A Corpus and Evaluation Framework for Deeper Understanding of Commonsense Stories
    • 89
    • PDF
    Ordinal Common-sense Inference
    • 65
    • PDF
    Men Also Like Shopping: Reducing Gender Bias Amplification using Corpus-level Constraints
    • 291
    • PDF
    Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering
    • 687
    • PDF
    A SICK cure for the evaluation of compositional distributional semantic models
    • 369
    • PDF
    A large annotated corpus for learning natural language inference
    • 1,646
    • Highly Influential
    • PDF
    Revisiting Visual Question Answering Baselines
    • 196
    • PDF