INDEX

Explanations

comparative sequence, self occupied

self-attention

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

ĳ

0.51

ersch

0.47

owitz

0.45

 마지막

0.45

 arginine

0.45

 Альбер

0.44

 তোমাদের

0.43

 спустя

0.43

amay

0.43

 Его

0.42

POSITIVE LOGITS

Multiply

0.47

Flu

0.44

民主

0.40

Uh

0.40

 formular

0.39

ുമോ

0.39

سین

0.38

'";

0.38

BBC

0.38

症

0.38

Activations Density 0.000%