INDEX

Explanations

code, scripting, files

The neuron is picking out tokens that contain internal dots—e.g. file names, class or object names, and version strings with “.” in them.

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

がない

-1.14

'-'

-0.98

³

-0.98

jand

-0.97

 schermata

-0.91

reason

-0.91

...')

-0.90

 مي‌

-0.90

'[

-0.89

 bouteilles

-0.89

POSITIVE LOGITS

png

1.13

をします

1.03

 aastal

1.00

واصل

0.98

exe

0.94

 "'";

0.93

مئ

0.93

 permitted

0.92

 }}$.

0.91

。”

0.91

Activations Density 0.224%