INDEX

Explanations

questions and requests

asking for specifics

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

 тобой

0.49

 тебе

0.47

 тобі

0.46

 তোমাকে

0.46

zeczytaj

0.45

 ćete

0.45

你要

0.43

你就

0.42

当你

0.41

 tengas

0.41

POSITIVE LOGITS

 Does

1.20

Does

1.15

 Could

1.06

 Would

1.05

 HELP

1.05

 Please

1.04

 wondering

1.02

 Help

0.98

請問

0.98

Would

0.97

Activations Density 0.434%