INDEX

Explanations

many

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 مرئيه

-1.19

 مشين

-0.96

Lähteet

-0.86

 OFDb

-0.82

脚注の使い方

-0.82

новниш

-0.78

expandindo

-0.77

 صوتيه

-0.76

 Roskov

-0.75

 prêtres

-0.75

POSITIVE LOGITS

 cases

1.07

cases

0.76

 ways

0.74

 places

0.70

 instances

0.64

 Fällen

0.64

but

0.61

 respects

0.58

 casos

0.56

 Cases

0.56

Activations Density 0.024%