INDEX

Explanations

precision experimental tests

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

 unless

-0.77

 stupid

-0.74

番茄

-0.73

 telescope

-0.69

ϋ

-0.68

 decent

-0.66

Hockey

-0.66

 clumsy

-0.66

 cannibal

-0.66

 Hockey

-0.65

POSITIVE LOGITS

 Precision

0.92

glück

0.91

Precision

0.91

EDM

0.90

Tests

0.86

EDM

0.85

 precision

0.85

CPT

0.83

 Ramsey

0.82

vergleich

0.82

Activations Density 0.027%