INDEX

Explanations

academic citations

The neuron fires on any square‐bracketed numeric reference or index, i.e. brackets immediately surrounding digits.

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

in

-1.73

 huit

-1.60

ᥬ

-1.48

pendentif

-1.41

 legyen

-1.34

ใน

-1.33

//=

-1.33

热闹

-1.29

 иссле

-1.28

Ee

-1.28

POSITIVE LOGITS

There

2.09

 уже

1.95

 they

1.92

 Those

1.91

Also

1.86

These

1.76

 There

1.75

 like

1.73

Those

1.69

 also

1.63

Activations Density 0.032%