INDEX

Explanations

error or cell names

The neuron fires on numeric literal tokens (especially floating‐point numbers).

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

-1.54

);

-1.52

-1.43

){

-1.34

ⅽ

-1.32

may

-1.29

-1.28

const

-1.27

-1.24

POSITIVE LOGITS

 Congrès

1.45

danke

1.34

牙齿

1.34

 排骨

1.27

majority

1.27

 Igual

1.26

釕

1.24

schen

1.22

لعاب

1.20

Technik

1.20

Activations Density 0.009%