INDEX

Explanations

chief followed by a role

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

});

-1.72

</h3>

-1.52

"""

-1.48

 LITER

-1.40

 nettsted

-1.39

𓆜

-1.36

`.

-1.34

 любом

-1.32

 ALREADY

-1.31

늄

-1.31

POSITIVE LOGITS

has

1.51

is

1.45

was

1.34

 became

1.31

 stated

1.30

 Often

1.27

 says

1.24

 gives

1.23

 becomes

1.23

 found

1.20

Activations Density 0.034%