INDEX

Explanations

conjunctions and connectors

arguments surrounding, prioritize public, breaking the mold

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

 вокруг

0.47

ブラシ

0.47

 COLL

0.44

 थोडे

0.44

 вроде

0.43

 overalls

0.43

書の

0.42

不太

0.42

 অন্যরা

0.42

的功能

0.41

POSITIVE LOGITS

 తి

0.49

 випад

0.47

 reden

0.46

ዔ

0.45

ارة

0.44

 Charakter

0.44

 erkannt

0.44

 erkennen

0.44

 reasons

0.43

 არს

0.43

Activations Density 0.001%