INDEX

Explanations

quantified items

model responses

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

');

0.45

.');

0.43

ymal

0.42

.')

0.41

'.

0.39

 renovations

0.39

getInt

0.38

 unbelievably

0.38

 छत्तीस

0.38

mal

0.37

POSITIVE LOGITS

 दिलचस्प

0.53

 دلچ

0.52

 সম্ভবত

0.48

 তুলনায়

0.47

ল্যান্ডের

0.45

<0x0C>

0.43

 vorg

0.43

PROGRESS

0.43

鱈

0.43

 ціка

0.42

Activations Density 0.023%