INDEX

Explanations

avoiding jargon and redundancy

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

 newData

2.05

trone

1.94

achment

1.91

legas

1.88

yesi

1.87

ulho

1.84

ạt

1.84

(°

1.82

yawa

1.82

arasi

1.82

POSITIVE LOGITS

 pitfalls

1.65

 altogether

1.43

 kosten

1.42

 obligatorio

1.41

 pozorn

1.40

 obvious

1.40

 disadvantages

1.39

 mandatory

1.37

例外

1.35

paying

1.35

Activations Density 0.073%