INDEX

Explanations

Ultimately, the "best"

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ロナ

0.88

 ہو۔

0.87

יו

0.86

นี้

0.84

หาย

0.83

ภา

0.82

 таких

0.81

เคย

0.81

山の

0.81

 อาจ

0.80

POSITIVE LOGITS

 mathematically

1.37

Daniels

1.24

 ornamented

1.21

 dismantled

1.20

 majest

1.19

 electrically

1.19

 mechanistic

1.18

 undertaking

1.17

 interestingly

1.17

 silico

1.16

Activations Density 0.000%