INDEX

Explanations

deal, derivatives, power, commodities

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

 nggak

-1.25

 lanterna

-1.23

 cowok

-1.20

 sorgfäl

-1.19

ꙁ

-1.18

アプリケーション

-1.15

 provinciale

-1.13

ὸν

-1.13

EndProject

-1.12

 hervorge

-1.10

POSITIVE LOGITS

бил

1.30

각

1.23

unächst

1.20

 çünkü

1.17

 Vereinigte

1.16

ваются

1.12

 verksamhet

1.09

 each

1.08

訖

1.07

 ит

1.06

Activations Density 0.079%