INDEX

Explanations

numbers

The neuron activates on numeric tokens (numbers, digits, and numeric symbols) 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

这

-1.22

 always

-1.15

これを

-1.11

 ポート

-1.10

名詞

-1.09

 パス

-1.08

这个

-1.07

 latter

-1.05

に

-1.04

 after

-1.03

POSITIVE LOGITS

 cuanto

1.16

感じで

1.13

為に

1.11

 فقط

1.05

my

1.05

ことも

1.03

lc

1.03

 éta

1.02

года

1.02

imanapun

1.02

Activations Density 0.068%