INDEX

Explanations

class initialization methods

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 사용하여

1.93

非常的

1.87

 위하여

1.86

 대하여

1.78

 می‌باشد

1.69

 organization

1.67

жён

1.64

 이용하여

1.62

하였습니다

1.61

 Optimization

1.60

POSITIVE LOGITS

litre

1.67

hota

1.56

hit

1.45

hra

1.44

hitting

1.43

tonne

1.42

lcii

1.39

 mlijeka

1.38

 பிரச்

1.38

obbly

1.38

Activations Density 0.130%