INDEX

Explanations

parenthetical phrases, citations, numbers

advice, disclaimers

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

Tarea

0.39

 clonal

0.39

 manually

0.38

𝗨

0.38

 Phaser

0.37

 prefect

0.36

体力

0.36

 doen

0.36

 TASK

0.36

 пе

0.35

POSITIVE LOGITS

erset

0.39

peter

0.38

regar

0.38

 narzęd

0.38

право

0.37

Peter

0.37

equality

0.37

ambio

0.36

roa

0.36

രായ

0.36

Activations Density 0.030%