INDEX

Explanations

entire groups or collections

The neuron fires on phrases emphasizing the completeness of something—e.g. “whole” or “entire” followed by a noun.

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

-1.44

 ragazza

-1.44

 toutefois

-1.34

will

-1.31

ff

-1.31

 Gesamt

-1.30

‟

-1.30

 trois

-1.30

 ebenfalls

-1.30

„

-1.29

POSITIVE LOGITS

1.75

 other

1.70

 that

1.67

 different

1.66

 these

1.55

 which

1.52

licante

1.46

)),

1.45

not

1.43

 erila

1.43

Activations Density 0.022%