INDEX

Explanations

specific concept identification

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ἢ

1.28

zość

1.22

?;

1.12

or

1.11

 QVector

1.10

ifrån

1.10

ή

1.09

?:

1.09

?—

1.08

 或者

1.06

POSITIVE LOGITS

ช

1.29

્

1.27

കേന്ദ്ര

1.25

Rib

1.21

Rib

1.21

Mush

1.20

部位

1.16

Ch

1.13

Kh

1.13

ม

1.11

Activations Density 0.000%