INDEX

Explanations

up and down

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

ptonshire

-0.57

ThroughAttribute

-0.57

 Octo

-0.55

NamedQuery

-0.55

LookAnd

-0.54

 Telex

-0.54

 Rayleigh

-0.54

oredCriteria

-0.54

ellants

-0.52

Eksterne

-0.52

POSITIVE LOGITS

 down

1.40

up

1.20

down

1.18

 Down

1.12

Down

1.09

 DOWN

1.06

Up

1.05

 downward

1.00

 upward

1.00

up

0.94

Activations Density 0.001%