INDEX

Explanations

model responses to wrong inputs

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

strange

1.27

🌴

1.27

 COLLECTION

1.26

Divine

1.25

Collection

1.25

 Lizard

1.24

 Lending

1.24

 Concerning

1.24

Collectible

1.22

 Breeding

1.22

POSITIVE LOGITS

 wipers

1.05

íoch

1.04

が進

0.90

ход

0.89

 נפ

0.87

σή

0.86

额

0.86

ud

0.85

is

0.85

áin

0.85

Activations Density 0.008%