INDEX

Explanations

introduction sections or descriptions

This neuron fires on numeric tokens (digits, measurements, route numbers, etc.).

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

 دقت

-0.84

訫

-0.83

 deres

-0.76

 Mungkin

-0.76

 termo

-0.75

 derro

-0.75

++)

-0.74

colgante

-0.73

芣

-0.72

یکن

-0.71

POSITIVE LOGITS

 abomination

0.87

 Divided

0.84

 Sperm

0.84

 الاج

0.84

 diva

0.80

 recie

0.79

機関

0.78

襬

0.78

 vervol

0.78

Laj

0.77

Activations Density 0.002%