INDEX

Explanations

processes and developments

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

 just

-1.45

 every

-1.13

 from

-1.09

 after

-1.08

out

-1.06

५

-1.05

 keeping

-1.05

 their

-1.05

 keep

-1.04

！」

-1.03

POSITIVE LOGITS

了出来

1.49

摟

1.41

or

1.41

 migli

1.39

 alebo

1.31

 mungkin

1.28

assemblée

1.28

 hoặc

1.26

may

1.25

อาจ

1.25

Activations Density 0.011%