INDEX

Explanations

numbers followed by punctuation

word problems with numbers

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

SERVIDOR

0.47

าส

0.46

菉

0.46

ToExp

0.45

Maar

0.44

Naj

0.43

会让

0.43

Algo

0.42

驽

0.42

ICreate

0.42

POSITIVE LOGITS

 found

0.54

 diminished

0.51

 seen

0.50

lob

0.47

0.45

 certain

0.45

 detected

0.45

 twice

0.44

0.43

Activations Density 0.001%