INDEX

Explanations

references to broadening or widening effects

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 cooling

-0.59

 AppComponent

-0.57

nloa

-0.54

下载附件

-0.52

cooling

-0.51

brica

-0.50

cektir

-0.50

])));

-0.50

Datuak

-0.49

 étag

-0.49

POSITIVE LOGITS

 widening

2.25

 widen

2.16

 widened

1.84

 broadening

1.56

 broaden

1.37

 broadened

1.21

Wid

1.20

wid

1.16

Wid

1.16

wid

1.05

Activations Density 0.003%