INDEX

Explanations

the fragment "Asp h", but also the word "asp" followed by a period

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 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

AEC

-0.67

 kteří

-0.66

"]]

-0.65

 Unito

-0.64

ã

-0.63

 caneca

-0.63

Ã

-0.63

föra

-0.63

Hek

-0.59

gons

-0.59

POSITIVE LOGITS

Asp

1.86

Asp

1.72

ASP

1.47

asp

1.32

asp

1.28

 aspi

1.27

 Aspire

1.25

aspir

1.23

 aspirin

1.22

 aspiration

1.20

Activations Density 0.013%