INDEX

Explanations

not pleasant

negating desirable qualities

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

 enough

0.59

 viktigt

0.58

 important

0.55

 even

0.55

 중요

0.54

 Important

0.52

 Enough

0.52

 viktig

0.52

 EVEN

0.52

 Even

0.51

POSITIVE LOGITS

 pleasant

0.87

pleasant

0.86

 приятно

0.86

 прият

0.80

прият

0.68

 bode

0.68

 agradable

0.66

 благоприят

0.60

 favorable

0.59

 Pleasant

0.59

Activations Density 0.090%