INDEX

Explanations

com

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ing

-0.60

-0.47

<eos>

-0.47

printStackTrace

-0.45

la

-0.45

es

-0.44

ING

-0.42

ed

-0.40

ه

-0.40

POSITIVE LOGITS

BeginContext

1.03

 itſelf

0.93

 themſelves

0.88

SourceChecksum

0.88

 doubtnut

0.87

 Majefty

0.87

 uſed

0.86

setVerticalGroup

0.86

 ―――――

0.85

DockStyle

0.85

Activations Density 0.266%