INDEX

Explanations

prepositional phrases indicating location or time

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

oba

1.36

entino

1.33

ayani

1.27

CIES

1.22

 (\%)

1.20

 recepción

1.10

^{*}}

1.08

まな

1.08

dotto

1.08

WARE

1.05

POSITIVE LOGITS

 suburban

2.61

 imper

2.48

 intermin

2.47

 precisely

2.47

 arcane

2.47

 Yale

2.43

 bizarre

2.43

 Harvard

2.40

 imposs

2.38

 Princeton

2.37

Activations Density 0.072%