INDEX

Explanations

mathematical theorems and results

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

hause

-1.05

 alcuni

-1.04

秋冬

-1.03

iamo

-1.02

ciled

-1.02

Ideally

-1.01

denk

-1.01

 temat

-1.01

samt

-1.01

ícias

-0.98

POSITIVE LOGITS

 because

1.28

our

1.23

 even

1.20

 there

1.15

 again

1.13

 nuevamente

1.11

if

1.11

 like

1.10

 remarkable

1.07

we

1.04

Activations Density 0.006%