INDEX

Explanations

Nobel Prize

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

excellence

2.12

 Nationwide

2.03

วัล

2.01

 Центра

1.98

ortis

1.95

wooden

1.90

eté

1.89

етки

1.88

トヨタ

1.87

<unused1768>

1.87

POSITIVE LOGITS

 bind

1.64

Binding

1.52

 চাহিয়া

1.51

load

1.47

 load

1.43

 intercamb

1.42

bind

1.40

 الاخت

1.35

携

1.34

മ്മ

1.33

Activations Density 0.017%