INDEX

Explanations

research and projects

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

 chutney

1.99

 pren

1.90

 kuat

1.88

 dicho

1.86

찹

1.83

 nett

1.83

 tribun

1.83

 kejadian

1.82

urence

1.81

 oración

1.81

POSITIVE LOGITS

gate

1.39

station

1.26

Gate

1.18

 phí

1.16

abandon

1.15

horses

1.15

team

1.13

Tracks

1.12

band

1.08

室

1.04

Activations Density 0.122%