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

rig

-0.29

rig

-0.25

Rig

-0.25

å¤±

-0.25

åºľ

-0.24

ì¡±

-0.24

åħīèį£

-0.24

ma

-0.24

ILITY

-0.24

çĽ¤

-0.24

POSITIVE LOGITS

relu

0.28

oten

0.28

á¸į

0.26

herits

0.26

enso

0.26

ous

0.26

orphic

0.25

è¿Ļä¹Īä¹ħ

0.25

è¿ŀèĥľ

0.25

oogle

0.24

Activations Density 0.886%

No Known Activations

This feature has no known activations.