INDEX

Explanations

guarantee you that

The neuron fires on words used in direct assurances or guarantees to the reader (e.g. “assure,” “guarantee,” “promise,” “trust me,” often coupled with “you” and “that”).

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

生素

-1.67

].

-1.57

 hunde

-1.55

__':

-1.41

НОСТИ

-1.34

 sorts

-1.32

 normales

-1.30

 générale

-1.30

 solids

-1.27

üz

-1.27

POSITIVE LOGITS

 there

2.73

There

1.55

 during

1.38

 that

1.34

there

1.29

 есть

1.29

コミック

1.27

 её

1.27

 this

1.26

 because

1.26

Activations Density 0.017%