INDEX

Explanations

technical keywords and phrases

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

」、「

-1.72

-1.57

-1.56

 ۱۱

-1.55

꯸

-1.52

Боль

-1.51

埽

-1.51

ться

-1.50

 ordained

-1.49

 edizione

-1.49

POSITIVE LOGITS

the

2.33

of

2.23

 their

2.02

our

1.80

 защото

1.71

 those

1.70

 these

1.66

</h2>

1.66

 such

1.55

ℱ

1.50

Activations Density 0.192%