INDEX

Explanations

Default class prefixes

This neuron primarily activates on occurrences of the identifier “Default” (e.g. class or type names prefixed with “Default”) in code.

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

to

-1.42

on

-1.41

↵↵↵

-1.36

if

-1.35

 when

-1.32

一刻

-1.31

“

-1.31

ться

-1.30

 porque

-1.30

POSITIVE LOGITS

 voyez

1.69

 maji

1.62

 soyez

1.51

 pensez

1.48

 pourriez

1.43

 mahd

1.42

 patienter

1.40

نور

1.39

ヶ

1.38

 trouvez

1.38

Activations Density 0.017%