INDEX

Explanations

problematic situations

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

𔘓

-1.42

ceda

-1.41

Vezi

-1.39

 játék

-1.38

🛢

-1.37

𓆉

-1.35

叕

-1.35

mittag

-1.34

 different

-1.34

 разные

-1.34

POSITIVE LOGITS

 then

1.92

1.49

1.48

1.44

 furthermore

1.40

for

1.38

lanan

1.37

if

1.34

1.28

 thereby

1.27

Activations Density 0.063%