INDEX

Explanations

The neuron activates on standalone numeric tokens (i.e. numbers or digits).

or if conditions

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

⬘

-1.09

 DELLE

-1.07

elcome

-1.06

 your

-1.05

%";

-1.05

our

-1.02

 Rasen

-1.02

 Poliz

-1.01

congrats

-1.00

姞

-1.00

POSITIVE LOGITS

or

1.66

if

1.40

 если

1.22

 will

1.18

 случае

1.09

 либо

1.02

 अगर

0.98

 якщо

0.96

各種

0.94

 wenn

0.93

Activations Density 0.007%