Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

Dataset: 80% C4 (Web Text) and 20% Python Code

Neuron 3961 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.1040. Min Act: -0.0001

Data Index: 218037 (C4 (Web Text))

Max Activating Token Index: 46

Text #1

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.1032. Min Act: -0.0001

Data Index: 66677 (C4 (Web Text))

Max Activating Token Index: 55

Text #2

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0977. Min Act: -0.0001

Data Index: 724702 (C4 (Web Text))

Max Activating Token Index: 40

Text #3

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0971. Min Act: -0.0001

Data Index: 395307 (C4 (Web Text))

Max Activating Token Index: 45

Text #4

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0929. Min Act: -0.0001

Data Index: 1541898 (Python Code)

Max Activating Token Index: 99

Text #5

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0961. Min Act: -0.0001

Data Index: 684980 (C4 (Web Text))

Max Activating Token Index: 39

Text #6

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0878. Min Act: -0.0001

Data Index: 450818 (C4 (Web Text))

Max Activating Token Index: 622

Text #7

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0945. Min Act: -0.0001

Data Index: 1498772 (Python Code)

Max Activating Token Index: 56

Text #8

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0930. Min Act: -0.0001

Data Index: 1587539 (Python Code)

Max Activating Token Index: 106

Text #9

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0894. Min Act: -0.0001

Data Index: 216823 (C4 (Web Text))

Max Activating Token Index: 796

Text #10

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0923. Min Act: -0.0001

Data Index: 1081005 (C4 (Web Text))

Max Activating Token Index: 49

Text #11

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0928. Min Act: -0.0001

Data Index: 1093582 (C4 (Web Text))

Max Activating Token Index: 695

Text #12

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0954. Min Act: -0.0001

Data Index: 1472759 (Python Code)

Max Activating Token Index: 64

Text #13

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0915. Min Act: -0.0001

Data Index: 778162 (C4 (Web Text))

Max Activating Token Index: 46

Text #14

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0885. Min Act: -0.0001

Data Index: 1385737 (Python Code)

Max Activating Token Index: 167

Text #15

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0874. Min Act: -0.0001

Data Index: 794915 (C4 (Web Text))

Max Activating Token Index: 19

Text #16

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0907. Min Act: -0.0001

Data Index: 1282848 (C4 (Web Text))

Max Activating Token Index: 49

Text #17

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0908. Min Act: -0.0001

Data Index: 65434 (C4 (Web Text))

Max Activating Token Index: 41

Text #18

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0907. Min Act: -0.0001

Data Index: 1615547 (Python Code)

Max Activating Token Index: 249

Text #19

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0889. Min Act: -0.0001

Data Index: 977979 (C4 (Web Text))

Max Activating Token Index: 46