INDEX

Explanations

our or all

The neuron activates on numerical tokens and numeric expressions (counts, measurements, prices, statistics) 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

 gehabt

-1.38

 weißem

-1.37

 vorigen

-1.32

ciendo

-1.32

 AGAIN

-1.31

 jenem

-1.31

äure

-1.30



-1.28

 besonder

-1.28

حيط

-1.27

POSITIVE LOGITS

of

1.91

 ícone

1.05

 spont

1.04

 быстро

1.01

こともあります

0.98

łop

0.98

in

0.97

無視

0.96

 bravo

0.94

 четыре

0.93

Activations Density 0.019%