INDEX

Explanations

Expression pattern, This invention, Introduction

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

 bicic

-1.20

 sirena

-1.18

 bici

-1.16

 cortina

-1.15

 krim

-1.14

 sucul

-1.14

 beginning

-1.13

 sentra

-1.10

 geografia

-1.08

 misto

-1.08

POSITIVE LOGITS

Don

1.04

 najlep

1.03

 gerekiyor

1.01

 beroperasi

0.95

 from

0.95

at

0.95

dtype

0.94

сього

0.91

inerario

0.91

ilant

0.89

Activations Density 0.001%