INDEX

Explanations

Matricide, Matrimonial, Monticello, FFA, Bethel

The neuron fires on occurrences of the substring “Jah” in words (e.g. names or terms containing “Jah”).

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

-2.86

ༀ

-2.77

–

-2.73

垕

-2.53

–

-2.50

$-$

-2.34

亖

-2.33

is

-2.31

部分が

-2.30

遢

-2.30

POSITIVE LOGITS

 selben

3.33

 Если

2.77

犛

2.58

 According

2.56

 ブログ

2.34

 example

2.31

 Then

2.30

rów

2.22

 Someone

2.19

 cortas

2.17

Activations Density 0.005%