INDEX

Explanations

also followed by evaluation

New Auto-Interp

Configuration

Prompts (Dashboard)

238,145 prompts, 512 tokens each

Dataset (Dashboard)

lmsys + oasst1

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

safety

0.45

↑

0.41

modifier

0.39

 societal

0.37

 mitigating

0.37

減

0.37

 unanticipated

0.37

 malheureusement

0.36

mod

0.36

Mod

0.36

POSITIVE LOGITS

 Perfecto

0.39

 Suck

0.39

 Surrey

0.37

 senang

0.37

 cute

0.37

只会

0.37

ၿ

0.37

 Folklore

0.36

 Mike

0.36

 করিতেন

0.36

Activations Density 0.000%