INDEX

Explanations

the era of the next

The neuron is a detector for the definite article “the.”

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

 that

-1.88

鉝

-1.63

 parfüm

-1.63

⃢

-1.61

ेमाल

-1.59

ཫ

-1.55

 histó

-1.55

 costruire

-1.54

"",

-1.53

 ..."

-1.52

POSITIVE LOGITS

 seperti

1.66

1.55

1.50

<h2>

1.44

1.40

1.38

 системы

1.37

 ваш

1.35

 dará

1.35

Activations Density 0.057%