INDEX

Explanations

self-reported data

The neuron detects numeric tokens (e.g. sample sizes, percentages, decimal values) 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

but

-1.31

那么

-1.30

lecht

-1.23

冶

-1.22

 ее

-1.20

⏮

-1.20

tión

-1.19

 infecciones

-1.16

 suoi

-1.15

евич

-1.14

POSITIVE LOGITS

 будущем

1.30

 said

1.29

 tenían

1.24

 крестьян

1.23

话说

1.22

一辈子

1.20

 künftig

1.20

 diventare

1.16

我说

1.13

 Informações

1.12

Activations Density 0.144%