INDEX

Explanations

91 followed by number or letter

The neuron fires on the digit “1” (i.e. occurrences of the character ‘1’ within numbers).

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

we

-2.03

躥

-2.00

-1.98

-1.95

 laden

-1.88

kingen

-1.87

 bahwa

-1.84

ه

-1.82

Filename

-1.77

 considérons

-1.75

POSITIVE LOGITS

橚

2.17

2.14

凩

2.11

\_

2.11

."

2.11

 byť

2.08

}$,

2.06

CELONA

2.03

一般的な

2.03

),

2.03

Activations Density 0.018%