INDEX

Explanations

reciting numbers

The neuron activates on numeric tokens—digit sequences representing counts, measurements, ages, etc.

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

 Entfer

-1.22

mässig

-1.18

and

-1.17

its

-1.13

安く

-1.07

向けの

-1.06

の準備

-1.04

ものである

-1.03

aniem

-1.02

itul

-1.01

POSITIVE LOGITS

 niitä

1.26

然

1.24

岁

1.24

了出去

1.23

 mondta

1.22



1.21

を

1.20

in

1.20

もありますが

1.20

 ибо

1.19

Activations Density 0.047%