word followed by context specifier

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

-1.16

 opportunities

-1.13

to

-1.08

 bulunur

-1.06

-1.05

”

-1.05

-1.03

 alterações

-1.02

 religieuses

-1.02

POSITIVE LOGITS

ִּ

1.38

 vpon

1.35

 pistolet

1.30

ַּ

1.28

Karakter

1.27

ּוֹ

1.26

FFECT

1.23

־ה

1.23

᾿

1.21

他們的

1.20

Activations Density 0.045%