INDEX

Explanations

average to mediocre quality

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 tablespoon

-0.94



-0.87

letka

-0.86

ӑ

-0.84

CLP

-0.83

LOUIS

-0.82

ificantly

-0.82

 equipamentos

-0.81

ὅ

-0.81

 ancora

-0.81

POSITIVE LOGITS

 mediocre

2.73

 average

2.69

 mediocrity

2.38

medio

2.33

 moderate

1.98

average

1.96

 ordinary

1.95

Medio

1.93

 medio

1.88

 Medio

1.87

Activations Density 0.050%