INDEX

Explanations

target

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

Life

1.24

cology

1.22

LP

1.22

LP

1.13

 профе

1.10

LS

1.10

GE

1.09

 Life

1.08

MA

1.08

 profes

1.08

POSITIVE LOGITS

 unrecognizable

1.33

 waitUntil

1.33

 rebounded

1.32

 TARGET

1.31

 overtook

1.30

 replenished

1.29

 expecting

1.29

 retribution

1.28

深夜

1.28

 recovering

1.27

Activations Density 0.001%

No Known Activations