INDEX

Explanations

both / two

The neuron activates on personal‐name tokens and date/number tokens—i.e. bits of biographical detail.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 celles

0.69

Those

0.68

those

0.68

 those

0.64

 Ceux

0.61

 Those

0.56

 mijn

0.56

 افرادی

0.56

getEntity

0.55

meme

0.55

POSITIVE LOGITS

两人

1.70

兩人

1.69

 juntos

1.46

二人

1.45

 তাঁদের

1.42

 இருவரும்

1.40

 loro

1.39

他们的

1.38

 their

1.35

两者

1.33

Activations Density 0.197%