INDEX

Explanations

Qualities, techniques, and effects

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

-1.81

用

-1.43

It

-1.43

對於

-1.35

UNTIME

-1.30

estroying

-1.29

对于

-1.29

是一個

-1.28

σαν

-1.23

ยว

-1.22

POSITIVE LOGITS

</h2>

1.57

</strong>

1.55

at

1.32

 choice

1.30

 perhaps

1.29

نجان

1.28

sen

1.27

</em>

1.26

 البعض

1.23

 THEOREM

1.22

Activations Density 0.045%