INDEX

Explanations

negative findings and criticism

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

ׇ

-1.51

 چربی

-1.38

 religieuses

-1.33

 ἐκ

-1.23

֩

-1.22

 الجلد

-1.13

CCIÓN

-1.12

 گیاه

-1.11

 εἰ

-1.09

 الشعر

-1.09

POSITIVE LOGITS

ּוֹ

2.28

ִּ

2.06

ַּ

2.03

וּ

2.03

־ה

1.80

ֶּ

1.73

ְּ

1.48

וֹ

1.44

nadequate

1.36

᾿

1.33

Activations Density 0.046%