INDEX

Explanations

oldest

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Frid

-0.09

 Ferd

-0.09

ifu

-0.09

 vintage

-0.09

 Antique

-0.09

 analogue

-0.09

 midnight

-0.09

orra

-0.08

Sas

-0.08

 antique

-0.08

POSITIVE LOGITS

 trick

0.15

 tricks

0.13

 profession

0.13

 Tricks

0.11

 professions

0.11

 Trick

0.10

 Profession

0.10

ali

0.10

marks

0.09

mpp

0.09

Activations Density 0.005%