INDEX

Explanations

eliminate or abolish

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

 hjæl

-0.96

 hjælpe

-0.84

𝙇

-0.79

ɩ

-0.77

دد

-0.75

DZ

-0.75

 hjælp

-0.75

enangan

-0.75

 høj

-0.75

TWO

-0.75

POSITIVE LOGITS

elimin

5.13

Elimin

5.00

 eliminate

4.88

 eliminated

4.84

 eliminating

4.59

 elimination

4.47

 Elimin

4.44

 eliminates

4.34

 Eliminate

4.25

 elimin

4.00

Activations Density 0.202%