INDEX

Explanations

conjunctions in multiple languages

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

 Василь

4.74

 يس

4.50

 Тимо

4.44

 Коло

4.42

 Михай

4.40

 يف

4.39

 Иванов

4.39

 يب

4.36

 Алек

4.36

 Сури

4.35

POSITIVE LOGITS

それに

0.32

更高的

0.23

amp

0.18

</h3>

0.16

<strong>

0.16

richten

0.14

</h2>

0.14

たとえば

0.12

機会

0.11

 satunya

0.09

Activations Density 0.048%