INDEX

Explanations

transformation into

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

 REALLY

1.72

 regimen

1.69

 priorities

1.66

 kasus

1.66

 mainstay

1.65

さんも

1.65

 dread

1.63

 SHOULD

1.63

 recommend

1.62

 excels

1.62

POSITIVE LOGITS

 그대로

0.95

を経て

0.95

 Crimea

0.94

 onCreateView

0.93

통령

0.90

 successivamente

0.89

 rilas

0.88

forming

0.88

tosecond

0.87

letscher

0.87

Activations Density 0.144%