Putting neural networks under the microscope

Researchers pinpoint the “neurons” in machine-learning systems that capture specific linguistic features during language-processing tasks.

Source: http://news.mit.edu/2019/neural-networks-nlp-microscope-0201