INDEX

Explanations

require ('

The neuron fires on occurrences of the CommonJS “require(…)” module-import calls.

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

if

-1.61

},

-1.59

to

-1.55

by

-1.49

all

-1.48

 from

-1.46

 more

-1.36

 just

-1.29

 even

-1.27

”,

-1.27

POSITIVE LOGITS

 facili

1.57

シャレ

1.50

provides

1.45

 rassemb

1.45

そも

1.44

 ingenu

1.41

 coûts

1.40

ほとん

1.40

 effe

1.40

雰囲

1.38

Activations Density 0.006%