INDEX

Explanations

HTML structure

The neuron is activated by numeric tokens (i.e. literal numbers). It essentially finds 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

for

-1.54

 from

-1.51

on

-1.31

to

-1.20

in

-1.09

—

-1.02

 with

-0.98

-0.94

of

-0.91

all

-0.91

POSITIVE LOGITS

 capucha

1.18

 PHILL

1.09

 znajduje

1.06

 taşlı

1.05

 praticamente

1.05

豎

1.04

 embarazo

1.03

 constituer

1.02

 estancias

1.02

 detayl

1.02

Activations Density 0.026%