INDEX

Explanations

No Explanations Found

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

rove

-0.27

Verifier

-0.27

ReadWrite

-0.25

Wyn

-0.25

çľŁ

-0.25

GROUND

-0.25

å¤ĸ

-0.24

ledge

-0.24

 verifier

-0.24

 Masc

-0.24

POSITIVE LOGITS

olith

0.30

 unpublished

0.27

bgcolor

0.26

ä¸ĢåĪĢ

0.25

hid

0.25

 Bashar

0.24

 sour

0.24

omid

0.24

proposal

0.24

åı²

0.23

Activations Density 0.004%

No Known Activations

This feature has no known activations.