INDEX

Explanations

direction and movement words

The neuron detects numeric measurement values (especially floating-point numbers and degrees/angles) in the text.

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

 чёр

-1.53

 investir

-1.41

/*"

-1.38

 csapat

-1.38

"#

-1.37

风雨

-1.34

掲示

-1.31

 OGSÅ

-1.30

 tornar

-1.27

南宫

-1.27

POSITIVE LOGITS

↵↵↵

1.68

主要的

1.58

1.37

’

1.35

怜

1.34

重要

1.33

making

1.31

 Gesamt

1.29

 etwas

1.27

主要

1.27

Activations Density 0.042%