Towards A Rigorous Science of Interpretable Machine Learning

On the reading lists

• #49 in Safety, alignment & interpretability (2026 edition)

Why it matters

Proposed a rigorous evaluation framework distinguishing application-grounded, human-grounded, and functionally grounded studies.

Condensed from the expert lists' notes.

The record

First released

2017 (from arXiv 1702.08608)

Citations

5,696 citations, as counted by Papers With Code (CC BY-SA 4.0)

Named by

both expert lists

Score

0.341 · computed rank 501 · consensus 0.17, must-read 0, citations percentile 0.83 (2010s · safety), awards 0, existing lists 0. Method