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

¢åįķ

-0.29

ost

-0.26

oster

-0.25

OST

-0.25

 anmeld

-0.24

åĲĦèĩª

-0.24

å®¡æŁ¥

-0.24

ott

-0.24

åĲĦ

-0.24

hp

-0.24

POSITIVE LOGITS

åģ¶

0.30

 INDIRECT

0.28

match

0.25

åħ¨å®¶

0.24

MATCH

0.24

/do

0.24

-wing

0.24

ansen

0.23

 Seymour

0.23

SplitOptions

0.23

Activations Density 0.061%

No Known Activations

This feature has no known activations.