INDEX

Explanations

characteristics and features

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 features

-1.11

features

-0.90

 characteristics

-0.88

 Features

-0.84

 FEATURES

-0.83

 traits

-0.77

characteristics

-0.73

Features

-0.72

 Characteristics

-0.71

 CHARACTERISTICS

-0.68

POSITIVE LOGITS

PerformLayout

0.64

heets

0.63

<bos>

0.53

 ziua

0.51

 Josephus

0.51

Tikang

0.51

quelize

0.50

heet

0.49

ContentAsync

0.49

 Cæsar

0.49

Activations Density 0.084%