INDEX

Explanations

code, math, structure

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

 despite

-1.26

-1.20

-1.19

on

-1.17

 that

-1.14

<!--[

-1.13

 from

-1.13

]]);

-1.12

now

-1.08

POSITIVE LOGITS

雰囲

1.48

 sommige

1.31

遛

1.21

綺麗な

1.21

きれいな

1.16

 verschillende

1.15

很多的

1.13

獗

1.13

Ragnar

1.13

色んな

1.10

Activations Density 0.003%