INDEX

Explanations

in a state or condition

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

the

-2.58

but

-1.82

 through

-1.70

at

-1.63

of

-1.62

Gå

-1.59

Få

-1.58

 throughout

-1.53

 sambil

-1.53

to

-1.49

POSITIVE LOGITS

↵

2.11

↵↵

1.77

 STILL

1.73

">

1.67

で一番

1.55

就是这样

1.52

 acabou

1.52

SEE

1.51

 utilizado

1.49

utiliz

1.48

Activations Density 0.034%