INDEX

Explanations

represented both

history, frozen, random, confidence

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

捷

0.54

Png

0.46

?!"

0.46

？」

0.46

ডিয়া

0.45

알

0.45

idirectional

0.44

दारों

0.44

仄

0.44

Threshold

0.44

POSITIVE LOGITS

 Belly

0.52

්

0.51

 latch

0.50

 teat

0.50

 Real

0.50

 fate

0.50

ilah

0.50

Ivy

0.48

Akh

0.48

Activations Density 0.001%