INDEX

Explanations

tournament

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

 therapy

-0.83

 treatment

-0.81

 Treatment

-0.76

treatment

-0.74

 TREATMENT

-0.72

 behandling

-0.69

 behandeling

-0.68

Treatment

-0.66

 terapia

-0.65

 tratamiento

-0.64

POSITIVE LOGITS

 <>",

0.98

addPreferredGap

0.66

 لينك

0.62

skosten

0.61

 fieldNum

0.61

aconda

0.60

 bugs

0.60

 PyTuple

0.59

клопе

0.58

StructEnd

0.58

Activations Density 0.129%