Most commonly used neuron explanation evaluation metrics fail two new sanity checks, and only Correlation, Cosine, AUPRC, F1-score, and IoU pass.
Understanding the role of individual units in a deep neural network
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Evaluating Neuron Explanations: A Unified Framework with Sanity Checks
Most commonly used neuron explanation evaluation metrics fail two new sanity checks, and only Correlation, Cosine, AUPRC, F1-score, and IoU pass.