INDEX

Explanations

start of something

This neuron detects mentions of getting off to a “start,” especially performance openings described with qualifiers like good, positive, flying, or fast.

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

</strong>

-1.73

 them

-1.69

</sub>

-1.58

 OGSÅ

-1.58

wanted

-1.50

</i>

-1.50

It

-1.48

‒

-1.48

</h3>

-1.47

するもの

-1.47

POSITIVE LOGITS

AllAfrica

1.49

McLaren

1.35

 tahun

1.35

on

1.34

 pueda

1.33

 sabía

1.32

леми

1.31

метров

1.30

瑧

1.29

糠

1.28

Activations Density 0.018%