INDEX

Explanations

excited to do things

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 услуг

-1.05

 related

-0.97

 entiendo

-0.96

 problem

-0.93

 чтения

-0.89

 Федерации

-0.86

use

-0.86

 نیاز

-0.83

can

-0.82

 ricerca

-0.81

POSITIVE LOGITS

 feels

1.19

 budeme

1.16

 RISERV

1.10

cense

1.08

we

1.08

our

1.07

Which

1.02

$\{

1.02

 excited

1.01

🏟

1.01

Activations Density 0.046%