INDEX

Explanations

comparisons and equivalences

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

鸶

-1.38

utenants

-1.30

</h2>

-1.27

 Gebruik

-1.26

ᮣ

-1.22

 écoute

-1.21

涤

-1.19

fications

-1.18

 Saltar

-1.16

 мечта

-1.16

POSITIVE LOGITS

甚至是

1.59

<i>

1.45

rozco

1.38

of

1.23

<b>

1.23

↵

1.23

 than

1.17

ământ

1.13

 निम्

1.12

 perceived

1.09

Activations Density 0.015%