INDEX

Explanations

righteousness and actions

The neuron activates on words and word parts related to “righteous” (e.g. righteous, righteousness, self-righteously).

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

am

-3.05

............

-2.83

…..

-2.75

.....

-2.72

….

-2.63

常用的

-2.61

-2.59

....

-2.56

''

-2.55

......

-2.53

POSITIVE LOGITS

驁

2.53

%@",

2.41

 становятся

2.34

ꦫ

2.30

ſu

2.25

 mannen

2.22

 sirena

2.17

 themſelves

2.16

 valla

2.11

 Répondre

2.06

Activations Density 0.002%