INDEX

Explanations

percentages or rates

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

क्

-1.04

ᄅ

-1.01

 jorge

-0.94

 PATTERN

-0.94

 тради

-0.94

ONU

-0.93

万美元

-0.92

ati

-0.92

textbackslash

-0.91

 fernando

-0.91

POSITIVE LOGITS

or

1.04

 when

1.01

GetAxis

0.96

 mengucap

0.96

galeria

0.96

 konkrét

0.95

 htons

0.93

igkeit

0.93

nimmt

0.93

jadikan

0.92

Activations Density 0.024%

No Known Activations