INDEX

Explanations

all complete

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

 várias

-1.04

 مباشرة

-0.90

 hierarchies

-0.87

lication

-0.86

onalds

-0.85

 способы

-0.84

">.

-0.82

xt

-0.82

(",",

-0.81

 cannons

-0.80

POSITIVE LOGITS

all

1.37

全部

1.06

 complete

1.06

Completion

1.04

 overall

1.02

 completion

1.01

 Completion

0.97

 rootReducer

0.94

 other

0.94

 camere

0.93

Activations Density 0.004%