INDEX

Explanations

variables and characteristics

The main thing this neuron does is find numeric tokens (literal numbers and numeric values) in the text.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 kawa

0.61

 placas

0.57

aceans

0.56

 salas

0.56

 choque

0.55

 precincts

0.55

 शिक्

0.54

 станови

0.54

 patrols

0.54

 अद्भ

0.54

POSITIVE LOGITS

字段

1.59

FieldName

1.48

fieldName

1.42

 columnName

1.41

fieldname

1.38

 field

1.38

 column

1.30

变量

1.30

 fields

1.29

 FIELD

1.28

Activations Density 0.104%