INDEX

Explanations

connecting list items

The neuron strongly activates on numeric tokens (numbers or measurements) 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

tisgarh

-1.36

ﬀ

-1.25

möglichen

-1.22

 głównie

-1.12

 }^{*}

-1.12

必要がある

-1.09

corer

-1.09

ectores

-1.07

 svých

-1.06

POSITIVE LOGITS

 there

2.80

 these

1.30

there

1.28

we

1.27

 they

1.23

そこは

1.21

some

1.20

個人的には

1.20

these

1.18

入り口

1.15

Activations Density 0.077%