INDEX

Explanations

easier to understand

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

 different

0.99

ต่างๆ

0.97

Are

0.95

different

0.89

不同的

0.89

Different

0.85

 differing

0.83

Ook

0.83

들이

0.82

Are

0.82

POSITIVE LOGITS

few

1.30

 pochi

1.24

 pocas

1.22

 pocos

1.16

 wenigen

1.13

 poucos

1.13

Few

1.11

Few

1.07

 reconnaît

1.02

 reconocer

1.01

Activations Density 0.205%