INDEX

Explanations

specific nouns and concepts

The neuron fires on content-bearing, often domain-specific nouns (especially longer or technical terms).

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

 they

-2.27

-1.54

ਫ਼

-1.48

Föld

-1.47

 their

-1.37

 glücklich

-1.37

 jejich

-1.35

ਖ਼

-1.34

 häufig

-1.33

 enfati

-1.31

POSITIVE LOGITS

 with

1.55

 without

1.48

 quella

1.28

 پارس

1.27

si

1.27

AndPassword

1.25

 inaprop

1.23

 ТА

1.23

 Venez

1.23

´

1.23

Activations Density 0.265%