INDEX

Explanations

Exactly. Pure. Nothing

This neuron fires on single-word, emphatic utterances—mostly adverbs or one-word sentences (e.g. “Ever.” “Seriously.” “Literally.”).

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

it

-1.99

mereka

-1.82

 ถ้า

-1.68

Suka

-1.65

Rumah

-1.64

Kontra

-1.63

Makan

-1.62

Edy

-1.61

menjadi

-1.58

 ideen

-1.57

POSITIVE LOGITS

also

1.82

],

1.64

糶

1.62

or

1.60

苙

1.59

许久

1.58

1.53

 年级

1.51

课

1.50

illoma

1.49

Activations Density 0.040%