INDEX

Explanations

code and technical specifications

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

Roman

-0.52

teady

-0.51

 Romans

-0.51

 Roman

-0.50

 wire

-0.48

 trenches

-0.46

tdi

-0.46

 Triassic

-0.45

Romans

-0.45

__))

-0.45

POSITIVE LOGITS

principalTable

0.85

ThroughAttribute

0.82

 linkovi

0.80

zzleHttp

0.75

UnknownFieldSet

0.71

DockStyle

0.66

Portale

0.66

AndEndTag

0.66

 initComponents

0.64

featureID

0.63

Activations Density 0.004%