INDEX

Explanations

whoever opponents

looking for strong success

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

 distinguir

0.49

ᴍ

0.46

λου

0.44

되었습니다

0.42

州的

0.42

ванные

0.42

 glic

0.42

 క్రికె

0.41

솎

0.40

 organizaciones

0.40

POSITIVE LOGITS

Whoever

0.43

}-${

0.42

ยอด

0.41

 opponents

0.40

 প্রাক্তন

0.40

梖

0.39

 Whoever

0.38

></

0.38

}\

0.38

 whoever

0.38

Activations Density 0.060%