INDEX

Explanations

adding items to collections

The neuron activates on numeric literals (especially floating‐point 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

CCIÓN

-1.54

 which

-1.43

but

-1.42

 mostly

-1.37

you

-1.29

 where

-1.28

 mainly

-1.27

してます

-1.24

kembangkan

-1.20

ري

-1.20

POSITIVE LOGITS

<td>

1.47

 это

1.36

Younger

1.33

⁷

1.32

麃

1.30

㈭

1.27

 Главная

1.27

 sechs

1.26

 steig

1.24

鉿

1.24

Activations Density 0.005%