INDEX

Explanations

numbers in parentheses

The neuron activates on numeric expressions (e.g. sample‐size or measurement values, often in parentheses).

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

尔夫

-1.45

 akzep

-1.38

ストン

-1.37

 autoridade

-1.37

 devenu

-1.36

я

-1.34

 utilizó

-1.33

作品

-1.31

ruzzo

-1.29

 Started

-1.28

POSITIVE LOGITS

and

1.60

\&

1.26

 сказала

1.24

انظر

1.16

 때문에

1.15

are

1.14

or

1.13

'],

1.11

was

1.09

.[

1.09

Activations Density 0.130%