INDEX

Explanations

likelihood of its design

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

 abhängig

1.06

Critic

1.02

critic

0.98

 liberals

0.97

宽松

0.97

issez

0.96

 tỉ

0.95

 idéal

0.93

Critics

0.93

 liberal

0.93

POSITIVE LOGITS

 belonged

1.31

 происхождения

1.08

belong

1.05

belongs

1.05

原來

1.02

 belongs

1.02

 принадлежа

1.02

原来

1.02

 guessed

1.01

 unidentified

0.99

Activations Density 0.613%