INDEX

Explanations

MR followed by specific terms

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

of

-3.33

ၜ

-3.17

some

-2.48

-2.42

for

-2.42

乀

-2.41

</sup>

-2.30

 "");

-2.28

ၟ

-2.28

ڧ

-2.25

POSITIVE LOGITS

ra

2.55

 entice

2.45

⃙

2.30

一旁

2.28

 információ

2.20

 кажется

2.17

 forego

2.14

re

2.13

战场

2.13

 mensen

2.09

Activations Density 0.008%