INDEX

Explanations

mathematical and system commands

The neuron strongly activates on tokens containing numeric characters (i.e. parts of numbers or digits).

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

支持

-1.03

 два

-0.99

 ме

-0.95

enced

-0.94

 кг

-0.93

lized

-0.93

hè

-0.93

де

-0.93

ølge

-0.92

со

-0.91

POSITIVE LOGITS

and

1.25

 until

0.97

to

0.96

 when

0.95

notif

0.89

無事に

0.84

購買

0.84

klady

0.84

ЧИТА

0.83

關係

0.83

Activations Density 0.064%