INDEX

Explanations

modules, risks, stories, engine

skill neurons

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

 colliding

0.48

רות

0.48

 annih

0.47

 anat

0.47

 pretending

0.46

 Elektrokh

0.46

𝟯

0.45

 Envelope

0.44

ળે

0.44

 repris

0.44

POSITIVE LOGITS

ip

0.52

结

0.49

 tyto

0.48

</tr>

0.47

ヽ

0.47

ürk

0.45

itie

0.44

 перший

0.44

ur

0.44

拿

0.44

Activations Density 0.000%