INDEX

Explanations

misunderstanding or misinterpretation

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

1.20

III

1.17

mg

1.11

1.06

 sôi

1.05

 terminate

1.03

not

1.03

 yaşayan

1.02

η

1.02

 شماره

1.01

POSITIVE LOGITS

splunk

1.40

𝐩

1.38

kep

1.38

राहुल

1.37

entrum

1.35

dzie

1.34

towards

1.31

ीण

1.31

tow

1.29

ه‌ها

1.29

Activations Density 0.184%