INDEX

Explanations

numbers indicating duration

The neuron strongly activates on numeric tokens (numbers and quantifiers).

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

of

-1.29

最多的

-1.27

 '\\;'

-1.22

それぞれ

-1.20

thirds

-1.15

が

-1.13

 odpowiada

-1.09

ÇÕES

-1.08

 nové

-1.06

Према

-1.06

POSITIVE LOGITS

 midt

1.05

 billig

1.03

 noin

1.03

了些

1.02

 ensuing

1.02

 HIER

1.00

 mereka

0.98

?!"

0.96

 colchón

0.94

äsident

0.93

Activations Density 0.136%