INDEX

Explanations

the word "proof"

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

 Burs

-0.85

 مؤرشف

-0.76

 وتسجيلات

-0.75

Personensuche

-0.74

 Pech

-0.73

 Digi

-0.73

Vgl

-0.71

...')

-0.71

 AppBundle

-0.71

 Orgel

-0.71

POSITIVE LOGITS

Proof

1.11

 PROOF

1.10

proof

1.09

 Proof

1.05

PROOF

1.01

 proof

0.91

 Proofs

0.87

 proofs

0.82

nyx

0.82

 bezeichneter

0.78

Activations Density 0.009%