INDEX

Explanations

situations where

The neuron is activated by standalone numeric tokens (numbers) 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

and

-2.86

or

-1.63

 konden

-1.54

&&

-1.52

zembro

-1.51

 maakte

-1.46

'';

-1.42

éndose

-1.40

 adverten

-1.40

zaamheid

-1.38

POSITIVE LOGITS

we

1.41

 everything

1.25

 Synag

1.24

Either

1.23

 there

1.23

は

1.22

 turísticos

1.22

 Meningkatkan

1.20

 chrétiens

1.20

他の

1.19

Activations Density 0.044%