INDEX

Explanations

down and up directionals

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

trine

-0.92

 próximos

-0.81

=='

-0.79

lī

-0.77

 acestea

-0.76

 nepiecieš

-0.75

aleta

-0.75

igail

-0.75

 automó

-0.73

 screenWidth

-0.73

POSITIVE LOGITS

 Down

1.16

Down

1.02

 DOWN

0.94

down

0.91

Up

0.88

DOWN

0.84

 cliente

0.81

缅

0.81

endiendo

0.81

疱

0.78

Activations Density 0.026%