INDEX

Explanations

The neuron detects forms of the word "said" when reporting speech or quotations.

instances of the word "said" and its variations, indicating dialogue or reported speech

New Auto-Interp

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

Negative Logits

jer

-0.86

ierrez

-0.77

Jer

-0.74

ambo

-0.73

omach

-0.72

renches

-0.71

igl

-0.71

stru

-0.70

aucas

-0.70

uras

-0.70

POSITIVE LOGITS

=#

0.82

 goodbye

0.69

 innocuous

0.68

lihood

0.63

ysis

0.62

 snippet

0.62

 rule

0.59

Yar

0.59

XP

0.58

 above

0.58

Activations Density 0.286%