INDEX

Explanations

you think you know

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

was

1.32

したのは

1.24

이었

1.20

 został

1.20

 tomó

1.19

was

1.16

となりました

1.15

 llegó

1.14

took

1.11

 mostró

1.10

POSITIVE LOGITS

 boast

1.32

 differ

1.21

owe

1.12

 know

1.10

 embody

1.07

 belong

1.06

 wanna

1.06

 regularly

1.05

 предложить

1.04

 need

1.03

Activations Density 0.274%