INDEX

Explanations

numbers and days

The neuron activates on numeric tokens (numbers, dates, measurements, decimals) 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

 even

-1.40

 darbo

-1.37

šte

-1.32

its

-1.26

 veiks

-1.25

 sexto

-1.21

 despite

-1.21

has

-1.20

 Neben

-1.20

 about

-1.17

POSITIVE LOGITS

就可以了

1.25

鸯

1.23

ודם

1.22

1.18

などが

1.17

icamente

1.16

՚

1.15

 FORCES

1.14

是非常

1.13

 مقالات

1.13

Activations Density 0.094%