Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 4439 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 2.2265. Min Range: -2.2265

Max Act: 2.2265. Min Act: -0.0000

Data Index: 1371587 (Python Code)

Max Activating Token Index: 46

Text #1

Max Range: 2.2265. Min Range: -2.2265

Max Act: 2.0747. Min Act: -0.0000

Data Index: 1186469 (C4 (Web Text))

Max Activating Token Index: 396

Text #2

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.9270. Min Act: -0.0000

Data Index: 942933 (C4 (Web Text))

Max Activating Token Index: 785

Text #3

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.9459. Min Act: -0.0000

Data Index: 1529284 (Python Code)

Max Activating Token Index: 661

Text #4

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.9218. Min Act: -0.0000

Data Index: 1026025 (C4 (Web Text))

Max Activating Token Index: 963

Text #5

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.8762. Min Act: -0.0000

Data Index: 746804 (C4 (Web Text))

Max Activating Token Index: 918

Text #6

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.7714. Min Act: -0.0000

Data Index: 1505783 (Python Code)

Max Activating Token Index: 4

Text #7

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.6390. Min Act: -0.0000

Data Index: 716535 (C4 (Web Text))

Max Activating Token Index: 943

Text #8

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.6116. Min Act: -0.0000

Data Index: 1139080 (C4 (Web Text))

Max Activating Token Index: 620

Text #9

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.6989. Min Act: -0.0000

Data Index: 1470207 (Python Code)

Max Activating Token Index: 79

Text #10

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.6955. Min Act: -0.0000

Data Index: 1045124 (C4 (Web Text))

Max Activating Token Index: 70

Text #11

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.5890. Min Act: -0.0000

Data Index: 1250966 (C4 (Web Text))

Max Activating Token Index: 675

Text #12

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.6354. Min Act: -0.0000

Data Index: 454500 (C4 (Web Text))

Max Activating Token Index: 599

Text #13

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.6893. Min Act: -0.0000

Data Index: 925775 (C4 (Web Text))

Max Activating Token Index: 853

Text #14

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.5702. Min Act: -0.0000

Data Index: 784611 (C4 (Web Text))

Max Activating Token Index: 691

Text #15

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.5002. Min Act: -0.0000

Data Index: 490669 (C4 (Web Text))

Max Activating Token Index: 138

Text #16

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.4960. Min Act: -0.0000

Data Index: 1195541 (C4 (Web Text))

Max Activating Token Index: 271

Text #17

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.5868. Min Act: -0.0000

Data Index: 1600973 (Python Code)

Max Activating Token Index: 697

Text #18

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.4892. Min Act: -0.0000

Data Index: 1618067 (Python Code)

Max Activating Token Index: 289

Text #19

Max Range: 2.2265. Min Range: -2.2265

Max Act: 1.5765. Min Act: -0.0000

Data Index: 883713 (C4 (Web Text))

Max Activating Token Index: 426