INDEX

Explanations

proper names and technical terms

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

 allemaal

-0.70

@[+][

-0.69

WireFormatLite

-0.68

MENAFN

-0.67

 متعلقه

-0.66

Havolalar

-0.62

URLException

-0.62

Hochspringen

-0.62

 Ανακτήθηκε

-0.60

antiate

-0.58

POSITIVE LOGITS

duo

1.05

Both

0.92

 pair

0.90

 Both

0.88

both

0.86

 beide

0.85

 twin

0.85

 both

0.84

 twins

0.81

 kedua

0.81

Activations Density 0.453%