INDEX

Explanations

"I" followed by discussion of recent events.

this neuron activates for just

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

Negative Logits

Palmarès

-0.54

InteropServices

-0.53

CppCodeGen

-0.52

℗

-0.51

はまだ

-0.51

AnimationsModule

-0.50

 már

-0.49

 isSet

-0.47

lapsingToolbar

-0.47

ImageContext

-0.47

POSITIVE LOGITS

ValueStyle

0.55

 tør

0.54

ational

0.53

pei

0.53

 hjäl

0.52

Batis

0.52

 مرئيه

0.51

 cure

0.51

Lur

0.51

kespea

0.50

Activations Density 0.597%