• Computer Science
  • Published in ArXiv 2019

ERASER: A Benchmark to Evaluate Rationalized NLP Models

@article{DeYoung2019ERASERAB,
  title={ERASER: A Benchmark to Evaluate Rationalized NLP Models},
  author={Jay DeYoung and Sarthak Jain and Nazneen Fatema Rajani and Eric Lehman and Caiming Xiong and Richard Socher and Byron C. Wallace},
  journal={ArXiv},
  year={2019},
  volume={abs/1911.03429}
}
State-of-the-art models in NLP are now predominantly based on deep neural networks that are generally opaque in terms of how they come to specific predictions. This limitation has led to increased interest in designing more interpretable deep models for NLP that can reveal the `reasoning' underlying model outputs. But work in this direction has been conducted on different datasets and tasks with correspondingly unique aims and metrics; this makes it difficult to track progress. We propose the… CONTINUE READING

References

Publications referenced by this paper.
SHOWING 1-10 OF 54 REFERENCES

Rationalizing Neural Predictions

VIEW 12 EXCERPTS
HIGHLY INFLUENTIAL

BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questions

VIEW 7 EXCERPTS
HIGHLY INFLUENTIAL

Rethinking Cooperative Rationalization: Introspective Extraction and Complement Control

VIEW 5 EXCERPTS
HIGHLY INFLUENTIAL

Neural Machine Translation by Jointly Learning to Align and Translate

VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL