INDEX

Explanations

names followed by titles or roles

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

 indiv

1.08

typen

1.05

 conceivable

1.05

geoType

1.05

 extraneous

1.04

 multiplicative

1.03

 problematic

1.02

 homogenous

1.02

 homogeneous

1.02

 covariant

1.01

POSITIVE LOGITS

1.42

1.41

’

1.35

Jr

1.23

،

1.19

 Productions

1.18

and

1.18

1.15

博士

1.10

 mengatakan

1.09

Activations Density 0.978%