INDEX

Explanations

the change is significant

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

ENERGY

-1.27

ないように

-1.25

on

-1.22

in

-1.15

 इस्तेमाल

-1.12



-1.11

派的

-1.09

 ngunit

-1.07

шают

-1.06

 Energy

-1.05

POSITIVE LOGITS

When

1.62

เมื่อ

1.45

Quando

1.41

 when

1.38

Cuando

1.37

If

1.37

ovács

1.29

 then

1.28

 partire

1.26

 lorsque

1.24

Activations Density 0.101%