INDEX

Explanations

code and comments

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

的上

-0.68

 skim

-0.68

 dissip

-0.67

mbps

-0.67

möglichkeiten

-0.67

WARN

-0.67

 proxies

-0.66

Mb

-0.66

Duncan

-0.66

 mercenaries

-0.65

POSITIVE LOGITS

pecabezas

0.73

 printable

0.70

mtext

0.69

ionic

0.67

SwitchCompat

0.66

Controlled

0.65

 heartbeat

0.65

眼鏡

0.64

onset

0.64

模式

0.64

Activations Density 0.062%