INDEX

Explanations

act as or like

The neuron strongly activates on occurrences of the root verb “act” (including its inflected forms like “acting”).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 Granted

0.68

 Bharati

0.60

 Equation

0.59

 ugljik

0.57

 kõr

0.57

 tiling

0.56

 tile

0.56

 khas

0.56

til

0.55

num

0.54

POSITIVE LOGITS

 акт

0.95

IVATE

0.93

ivating

0.81

Acts

0.80

act

0.79

upuncture

0.78

act

0.77

ively

0.76

Acting

0.76

Act

0.75

Activations Density 0.161%