INDEX

Explanations

place and food

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

eteria

-0.06

éĹĬ

-0.06

éĤ

-0.06

 irregular

-0.06

ental

-0.05

 extension

-0.05

Ã©ri

-0.05

_extension

-0.05

æĸ¹æŃ£

-0.05

 Komb

-0.05

POSITIVE LOGITS

 consum

0.06

 Thorn

0.06

bol

0.06

 cera

0.06

fg

0.06

 vows

0.06

 flattering

0.05

 intoxic

0.05

åħ¨æµģç¨ĭ

0.05

æķ´ä¸ªè¿ĩç¨ĭ

0.05

Activations Density 0.005%

No Known Activations