INDEX

Explanations

actions like playing or singing

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

zie

-0.09

ÛĮÙĨ

-0.09

hap

-0.08

 Paige

-0.08

ev

-0.08

iesel

-0.08

EGA

-0.08

-ev

-0.08

tos

-0.08

bere

-0.08

POSITIVE LOGITS

 Ð²Ð¾Ð·

0.11

 Accord

0.11

 boxing

0.10

_DER

0.10

 Invisible

0.10

(po

0.10

 closing

0.09

 sple

0.09

acin

0.09

ardin

0.09

Activations Density 0.031%

No Known Activations