INDEX

Explanations

assess, process, and improve

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

があり

-1.41

 poiché

-1.41

ですが

-1.37

Õ

-1.34

そのため

-1.29

׳

-1.29

ʻ

-1.27

がある

-1.27

 inay

-1.25

なので

-1.23

POSITIVE LOGITS

and

2.80

or

2.13

 казалось

1.42

},

1.35

bzw

1.32

 albeit

1.29

not

1.26

 With

1.21

 During

1.20

はな

1.18

Activations Density 0.050%