INDEX

Explanations

optimal

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

 optimal

-1.27

optimal

-1.27

Optimal

-1.21

 Optimal

-1.18

 optimally

-1.14

 optimum

-1.13

 automatiques

-1.05

 démocr

-1.03

 définiti

-1.00

 vectorielles

-0.98

POSITIVE LOGITS

ized

0.76

ization

0.70

izing

0.60

use

0.59

ised

0.58

ize

0.56

 care

0.54

 part

0.53

fra

0.52

 press

0.52

Activations Density 0.093%