INDEX

Explanations

and followed by various words

(and what)

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

kern

0.46

 இந்நிலையில்

0.44

asign

0.43

 Karnataka

0.42

 Farmington

0.41

iles

0.40

 Lebanon

0.40

 ہندوستان

0.40

brahim

0.39

ových

0.39

POSITIVE LOGITS

!).

0.61

!),

0.61

 oczywiście

0.60

presumably

0.59

 übrigens

0.59

 кстати

0.58

!)

0.57

прочем

0.57

 presumably

0.54

もちろん

0.52

Activations Density 0.171%