INDEX

Explanations

George Washington University

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

 Verpackung

-0.78

pulumi

-0.78

coupe

-0.77

 triom

-0.76

 noires

-0.75

vių

-0.75

UESDAY

-0.74

🪀

-0.73

WAG

-0.73

 trein

-0.71

POSITIVE LOGITS

GW

0.79

ɤ

0.71

ब

0.71

let

0.71

 verliert

0.70

ILE

0.68

 nationally

0.67

 lose

0.67

 without

0.66

 ready

0.65

Activations Density 0.009%