INDEX

Explanations

asking which to choose

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

ণিজ

2.12

лып

1.98

<blockquote>

1.82

IPTS

1.73

fim

1.72

1.71

்ட

1.69

ција

1.67

വസ്ഥ

1.63

ணுவ

1.61

POSITIVE LOGITS

 reprend

2.56

蜓

2.52

 thoroughly

2.47

Bxh

2.46

🏽

2.44

 violently

2.43

 ainfi

2.42

 fasten

2.42

Qxh

2.41

yǐ

2.40

Activations Density 0.277%