INDEX

Explanations

well-regarded or well-established

well-known and regarded

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

 major

0.63

流行

0.62

常用

0.59

常用的

0.59

 popular

0.59

 common

0.57

 iconic

0.55

 commonplace

0.55

 commonly

0.54

定番

0.54

POSITIVE LOGITS

regarded

0.91

 regarded

0.86

respected

0.82

 RESPECT

0.73

 respected

0.73

 regard

0.70

respect

0.69

Respect

0.69

 respect

0.68

 Respect

0.66

Activations Density 0.040%