INDEX

Explanations

lists of alternatives or conditions

The neuron fires most strongly on degree‐ or quantity‐expressing terms (e.g. superlatives like “best,” quantifiers like “few,” and duration units like “days” or “months”).

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

and

-1.53

such

-1.50

⅝

-1.46

to

-1.37

-1.36

in

-1.30

SAY

-1.29

⅜

-1.29

 компаний

-1.28

だけでなく

-1.27

POSITIVE LOGITS

tså

1.51

 samarbeid

1.45

 decorativas

1.43

ظمة

1.41

satunya

1.40

 when

1.39

 motivadoras

1.38

 peinados

1.38

 voulez

1.38

 elegancia

1.37

Activations Density 0.264%