INDEX

Explanations

app =

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

Miscellaneous

1.28

方面的

1.21

 аспек

1.18

ج

1.16

ข

1.16

ข

1.14

BAT

1.14

inguistic

1.09

 Letters

1.09

Kep

1.09

POSITIVE LOGITS

 έχει

1.85

 έχουν

1.77

 είναι

1.71

 diffère

1.70

 μπορούν

1.65

实际上

1.62

 geralmente

1.61

 plupart

1.60

 maioria

1.60

 seule

1.59

Activations Density 0.141%