INDEX

Explanations

mixed text types

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ura

-0.28

kernel

-0.26

Assistant

-0.25

URA

-0.25

å¸Ĥå§Ķå¸¸å§Ķ

-0.24

andez

-0.24

 skirts

-0.24

Ã¸d

-0.23

gay

-0.23

UA

-0.23

POSITIVE LOGITS

éĢĤåĲĪèĩªå·±

0.27

{})

0.27

mmo

0.26

è¿½è¸ª

0.26

Interaction

0.26

å¯»

0.25

åħ¬åħ³

0.25

 programmed

0.25

Psi

0.25

 Encounter

0.25

Activations Density 0.002%