INDEX

Explanations

-ation / -ization / -ating endings

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

liness

-0.09

eliness

-0.08

ings

-0.08

ners

-0.08

tres

-0.08

uko

-0.08

inges

-0.08

Ãªtre

-0.08

yk

-0.07

yth

-0.07

POSITIVE LOGITS

ated

0.22

ized

0.20

ation

0.19

ization

0.18

ating

0.18

izing

0.17

ate

0.16

isation

0.14

ised

0.14

ationToken

0.13

Activations Density 0.700%