INDEX

Explanations

citations and metadata

This neuron responds to document section headings and metadata labels (e.g. “Abstract,” “Aim,” “Skill level,” “Date,” “Funding information,” 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

 However

-1.53

бливості

-1.48

żenia

-1.45

撬

-1.42

₰

-1.37

Vocabulary

-1.33

</i>

-1.33



-1.30

ぼう

-1.30

ňuje

-1.28

POSITIVE LOGITS

1.91

それは

1.72

Any

1.72

With

1.70

Rather

1.65

&&

1.63

From

1.62

 любых

1.60

Here

1.57

Some

1.57

Activations Density 0.004%