Analyzing machine-learned representations: A natural language case study

@article{Dasgupta2020AnalyzingMR,
  title={Analyzing machine-learned representations: A natural language case study},
  author={Ishita Dasgupta and Demi Guo and S. Gershman and Noah D. Goodman},
  journal={Cognitive science},
  year={2020},
  volume={44 12},
  pages={
          e12925
        }
}
As modern deep networks become more complex, and get closer to human-like capabilities in certain domains, the question arises as to how the representations and decision rules they learn compare to the ones in humans. In this work, we study representations of sentences in one such artificial system for natural language processing. We first present a diagnostic test dataset to examine the degree of abstract composable structure represented. Analyzing performance on these diagnostic tests… Expand
3 Citations
Measuring Systematic Generalization in Neural Proof Generation with Transformers
  • 5
  • PDF
Computational Psychiatry for Computers

References

SHOWING 1-10 OF 83 REFERENCES
Human few-shot learning of compositional instructions
  • 36
  • PDF
Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference
  • 253
  • PDF
Relational inductive biases, deep learning, and graph networks
  • 973
  • PDF
Visualizing and Understanding Neural Models in NLP
  • 418
  • PDF
Building machines that learn and think like people
  • 1,122
  • PDF
...
1
2
3
4
5
...