INDEX

Explanations

type definitions and usage

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

 Tugend

-0.75

sified

-0.74

절

-0.73

 Neustadt

-0.73

оборот

-0.73

 Kassel

-0.73

where

-0.73

criteria

-0.72

giu

-0.72

фигу

-0.72

POSITIVE LOGITS

ServiceTest

0.81

 fundamental

0.80

Doge

0.74

 Erase

0.74

 underlying

0.73

áklad

0.72

んびり

0.71

 unmodified

0.70

的中

0.70

 base

0.69

Activations Density 0.031%