INDEX

Explanations

generation, creation, construction

The neuron fires on nouns naming processes or actions (especially nominalizations ending in “-ion” or “-ment,” e.g. generation, production, cultivation, creation).

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

 they

-2.17

 other

-1.80

値下げ

-1.61

 there

-1.57

 help

-1.51

 give

-1.47

ी

-1.45

 found

-1.42

to

-1.41

 take

-1.40

POSITIVE LOGITS

and

1.78

توضی

1.72

 étrang

1.63

 célè

1.58

 thậm

1.57

können

1.57

richtig

1.55

 Kläger

1.54

 fidé

1.53

 réinitial

1.53

Activations Density 0.323%