INDEX

Explanations

exclamations and direct address

The neuron flags tokens containing the letter sequence “no” (e.g. “no,” “know,” 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

 will

-1.72

who

-1.69

 whose

-1.62

 Those

-1.54

 using

-1.53

 those

-1.40

 usually

-1.33

 because

-1.32

one

-1.32

 which

-1.30

POSITIVE LOGITS

 vielen

1.43

 internacionales

1.38

 valla

1.34

FullPath

1.34

 malgré

1.30

糌

1.29

 своим

1.29

 accentuated

1.28



1.27

меры

1.26

Activations Density 0.028%