INDEX

Explanations

physics and scientific terms

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

use

-1.43

ing

-1.37

 påver

-1.17

-1.16

-1.15

 obrá

-1.13

be

-1.12

 création

-1.09

および

-1.06

 udě

-1.06

POSITIVE LOGITS

ugais

1.54

 かな

1.31

anthes

1.29

ممنون

1.27

ulturally

1.24

adell

1.23

 fracaso

1.23

 たち

1.22

<0xD3>

1.21

étisme

1.17

Activations Density 0.052%