INDEX

Explanations

extra identifier, vulnerability to

also, let me know

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

0.48

0.47

engkap

0.46

umber

0.44

 그다

0.44

0.42

ajos

0.41

lihat

0.41

诋

0.41

0.40

POSITIVE LOGITS

 Textiles

0.50

 harán

0.50

 Physically

0.49

эри

0.48

 quarantine

0.47

 Свет

0.47

 SWAT

0.47

ରି

0.46

 Textile

0.45

 blanch

0.45

Activations Density 0.001%