INDEX

Explanations

okay agreement or acknowledgement

This neuron fires on informal discourse markers or interjections (e.g. “ok,” “so,” “wow”) that kick off a reply.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

㉓

-1.62

?”

-1.56

 reconocido

-1.55

躂

-1.48

</h1>

-1.47

 TIME

-1.45

NEW

-1.45

 usein

-1.44

╿

-1.42

 帅气

-1.41

POSITIVE LOGITS

vieron

1.41

They

1.33

 jaguar

1.27

 funer

1.27

៚

1.26

 положении

1.24

 étoit

1.24

中略

1.23

ـ

1.22

 sachant

1.21

Activations Density 0.005%