INDEX

Explanations

++) followed by {

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

return

-1.84

 return

-1.70

 then

-1.64

 even

-1.52

or

-1.41

 returning

-1.35

so

-1.33

 Even

-1.24

but

-1.21

 returned

-1.21

POSITIVE LOGITS

 CONSTANT

1.13

whenever

1.12

quando

1.10

 magnific

1.07

justed

1.07

 strikingly

1.06

饸

1.06

눔

1.04

 RESPONSE

1.02

gefügt

1.02

Activations Density 0.013%