INDEX

Explanations

not followed by denial or qualification

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

3.34

er

3.17

ه

2.81

ে

2.64

iš

2.56

śni

2.55

aia

2.51

શી

2.46

רץ

2.43

udius

2.41

POSITIVE LOGITS

wreck

3.04

@"

3.02

PubKey

2.90

rocket

2.72

 οποίο

2.67

shuffle

2.64

 MAKEINTRESOURCE

2.64

!/"

2.63

ることが

2.63

की

2.61

Activations Density 0.059%