INDEX

Explanations

uh oh aha ha

The neuron activates on interjections and vocalized filler sounds (e.g. “uh,” “oh,” “aha,” onomatopoeic grunts, etc.).

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

to

-2.33

我一

-1.74

⸨

-1.62

躂

-1.60

</h5>

-1.59

 शुर

-1.54

-1.52

咴

-1.50

sandalias

-1.48

the

-1.47

POSITIVE LOGITS



1.75

们

1.67

이고

1.58

––

1.50

crania

1.47

 lyon

1.45

 ecuador

1.45

 プレミアム

1.45

 marseille

1.45

癱

1.43

Activations Density 0.005%