INDEX

Explanations

Technical and code-related texts

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 electric

-0.36

 electronically

-0.31

 installer

-0.31

 electrical

-0.30

 Electrical

-0.28

esson

-0.28

 Electric

-0.28

Electric

-0.28

 tÃ©lÃ©

-0.27

å¦ĤæľŁ

-0.27

POSITIVE LOGITS

æ¨¡å¼ı

0.28

ogi

0.28

chos

0.27

æİĴ

0.26

ABC

0.25

æĹĭ

0.25

è¿Ĳä½ľ

0.25

æĹĭè½¬

0.24

éļĶ

0.24

 multim

0.24

Activations Density 0.101%