INDEX

Explanations

data types followed by . or {

The neuron fires on mentions of core JavaScript built-ins—especially the global objects like JSON and Object.

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

for

-1.86

を

-1.77

 then

-1.70

are

-1.61

 each

-1.56

 república

-1.52

 other

-1.52

！！！

-1.52

 with

-1.51

！(

-1.49

POSITIVE LOGITS

 alho

1.59

 rassemb

1.56

ウチ

1.55

’

1.55

%;

1.52

偸

1.52

 They

1.50

 welchen

1.48

為に

1.48

を作って

1.45

Activations Density 0.007%