INDEX

Explanations

rules of or on

This neuron detects occurrences of the word “rule” (or its plural “rules”) in text.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

xvii

-1.41

であった

-1.39

itário

-1.31

师兄

-1.31

uttgart

-1.30

rieron

-1.30

-1.29

cómo

-1.29

 manera

-1.28

 озеро

-1.27

POSITIVE LOGITS

ぼ

1.82

 lunedì

1.69

as

1.69

dr

1.56

to

1.55

蓠

1.53

Rule

1.53

 beginning

1.52

 mieć

1.52

”;

1.50

Activations Density 0.028%