INDEX

Explanations

data structure names

The neuron is triggered by standalone programming identifiers—variable, function, class, and other code symbol names.

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

and

-1.47

 autres

-1.46

 voient

-1.38

 veulent

-1.34

 réunions

-1.29

 Univer

-1.28

 outras

-1.21

 erfolgreichen

-1.20

 proposent

-1.20

 séjours

-1.19

POSITIVE LOGITS

 When

1.38

zipCode

1.34

ILog

1.33

 cuando

1.27

fileId

1.24

給

1.24

 Cuando

1.23

 særlig

1.20

 човек

1.20

 llegamos

1.20

Activations Density 0.071%