INDEX

Explanations

data collection

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 collected

-1.74

 gathered

-1.52

collected

-1.46

 conducted

-1.44

conducted

-1.24

 held

-1.21

 carried

-1.19

 Collected

-1.13

 nahilalakip

-1.09

 organisée

-1.06

POSITIVE LOGITS

the

0.80

by

0.65

any

0.65

up

0.65

 onto

0.65

 next

0.64

0.60

all

0.59

 some

0.59

 these

0.59

Activations Density 0.076%