INDEX

Explanations

ranking or position

ordinal numbers and positions

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

 second

0.85

 вто

0.78

 segundo

0.73

 birinci

0.73

 third

0.73

second

0.71

 segunda

0.71

 thứ

0.71

 zweiten

0.71

SECOND

0.71

POSITIVE LOGITS

 spot

0.64

overall

0.58

position

0.56

 position

0.54

 overall

0.54

spot

0.49

 finisher

0.48

Overall

0.48

 Overall

0.47

 spots

0.47

Activations Density 0.043%