INDEX

Explanations

code structure and syntax

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

and

-1.30

for

-1.26

吅

-1.19

龶

-1.17

 leren

-1.13

၅

-1.07

 meille

-1.05

Often

-1.05

 pengertian

-1.05

 démocr

-1.04

POSITIVE LOGITS

 laborious

1.27

 struggles

1.24

𝅥

1.16

 Ordner

1.15

řád

1.14

 своем

1.13

 αὐτ

1.13

relaxed

1.13

 τὸν

1.13

 imprimé

1.11

Activations Density 0.030%