INDEX

Explanations

names of saints or religious places

The neuron activates on proper names of churches, temples, and other religious institutions.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 manquer

-1.73



-1.69

Exemple

-1.69

こともあります

-1.58

 couver

-1.57

㩒

-1.55

 ſont

-1.53

 incendie

-1.52

从来没有

-1.50

 demasi

-1.49

POSITIVE LOGITS

in

2.02

for

1.66

not

1.54

 when

1.52

 without

1.50

by

1.49

 where

1.49

For

1.48

we

1.45

～！

1.40

Activations Density 0.032%