INDEX

Explanations

calculation and updating

calculation process

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 постепенно

0.42

 propelling

0.42

 gradually

0.41

渐渐

0.41

 آہستہ

0.41

 спри

0.41

iddag

0.41

த்தா

0.41

हिणी

0.41

 कामयाबी

0.40

POSITIVE LOGITS

Calculate

0.46

Bl

0.45

 erneut

0.45

計算

0.42

次

0.41

Everything

0.41

 neue

0.40

Computing

0.40

Update

0.40

WITH

0.40

Activations Density 0.001%