INDEX

Explanations

punching bag

assessing good or poor outcomes

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

ות

0.50

Update

0.46

Mid

0.45

皞

0.45

Mystery

0.45

Stone

0.44

 मानचित्र

0.44

Solution

0.44

Black

0.43

Reporter

0.43

POSITIVE LOGITS

所得

0.42

 Möglichkeiten

0.40

 scales

0.40

 suspensions

0.39

 sunken

0.39

的就是

0.39

 gemeinsamen

0.39

斥

0.39

 acted

0.39

ંબ

0.38

Activations Density 0.000%