INDEX

Explanations

categories for people and media

This neuron activates on two-word descriptive noun phrases (an adjective or modifier followed by a noun), like “lowland forests,” “intermittent freshwater marshes,” or “heated leather seats.”

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

為に

-1.48

-1.46

as

-1.41

ようです

-1.40

-1.37

ところに

-1.36

 There

-1.35

 Both

-1.35

ための

-1.34

POSITIVE LOGITS

<bos>

1.96

がありますが

1.82

ꟷ

1.69

ジナル

1.63

brtc

1.61

Manfaat

1.56

や

1.55

などが

1.52

もあって

1.52

があるので

1.50

Activations Density 0.035%