INDEX

Explanations

steps in technical processes

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

åį«

-0.25

çļĦåĵģçīĮ

-0.24

ipy

-0.23

 medial

-0.23

éĶ¡

-0.23

ç»ħ

-0.22

rst

-0.22

è¡Ľ

-0.21

ä¹ĭå®¶

-0.21

 Brand

-0.21

POSITIVE LOGITS

éĺ¶æ®µ

0.26

 preced

0.25

åĴĮæĬĢæľ¯

0.24

åıĳå±ķéĺ¶æ®µ

0.23

æľ¬èº«å°±æĺ¯

0.23

æľ¬èº«å°±

0.23

åĪ°ä½į

0.23

atte

0.23

 Stage

0.23

æľ¬èº«

0.23

Activations Density 0.566%