INDEX

Explanations

sub

The neuron fires on occurrences of the prefix “sub” (i.e. words beginning with “sub-”).

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

يقة

0.64

 Cynthia

0.59

ية

0.58

 नजदीक

0.57

eas

0.56

acos

0.55

 Patti

0.55

 straightened

0.55

arrest

0.55

 Fleetwood

0.55

POSITIVE LOGITS

 суб

1.40

Sub

1.38

Sub

1.38

SUB

1.35

sub

1.34

sub

1.31

SUB

1.31

getSub

1.20

サブ

1.16

 subi

1.16

Activations Density 0.117%