INDEX

Explanations

currency, discounts, and numbers

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

 flavor

-0.98

 flavorful

-0.89

 specializing

-0.87

ardized

-0.87

 flavored

-0.87

 Flavor

-0.86

 colorful

-0.82

 uchun

-0.81

urized

-0.81

 behavior

-0.81

POSITIVE LOGITS

£

1.55

 reductions

1.25

Reduced

1.15

(£

1.14

 réduire

1.13

-£

1.11

£

1.10

 Argos

1.05

Tes

1.04

reduced

1.01

Activations Density 0.034%