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

èĥĮæĻ¯

-0.28

 NUnit

-0.27

éĢĤå½ĵçļĦ

-0.27

Ð»ÐµÐ¼

-0.27

éĢĤå½ĵ

-0.27

 studs

-0.26

èĬ¯çīĩ

-0.26

ciÃ³

-0.25

æľĢéĢĤåĲĪ

-0.24

åı·çłģ

-0.24

POSITIVE LOGITS

ignant

0.29

ual

0.27

 life

0.26

Cl

0.25

 intern

0.25

 resolution

0.24

 duty

0.24

resolution

0.24

__)↵↵

0.24

aga

0.24

Activations Density 0.009%

No Known Activations

This feature has no known activations.