INDEX

Explanations

mathematical equations

The neuron activates on numeric tokens—especially multi‐digit numbers and decimal values.

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

Про

1.38

part

1.36

make

1.34

for

1.34

color

1.34

person

1.33

match

1.33

not

1.30

return

1.28

minor

1.27

POSITIVE LOGITS

1.63

£

1.55

 Ihre

1.55

 Jika

1.52

 Ethyl

1.52

HEK

1.52

 Lebih

1.52

 Verlag

1.50

 June

1.49

USS

1.46

Activations Density 0.239%