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

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

Neuron 3128 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.6932. Min Range: -2.6932

Max Act: 2.6932. Min Act: -0.0000

Data Index: 1022232 (C4 (Web Text))

Max Activating Token Index: 485

Text #1

Max Range: 2.6932. Min Range: -2.6932

Max Act: 2.2726. Min Act: -0.0000

Data Index: 1318515 (C4 (Web Text))

Max Activating Token Index: 230

Text #2

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.9967. Min Act: -0.0000

Data Index: 981872 (C4 (Web Text))

Max Activating Token Index: 796

Text #3

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.8609. Min Act: -0.0000

Data Index: 1293971 (C4 (Web Text))

Max Activating Token Index: 43

Text #4

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.8595. Min Act: -0.0000

Data Index: 759639 (C4 (Web Text))

Max Activating Token Index: 755

Text #5

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.8087. Min Act: -0.0000

Data Index: 198295 (C4 (Web Text))

Max Activating Token Index: 185

Text #6

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.7592. Min Act: -0.0000

Data Index: 1468836 (Python Code)

Max Activating Token Index: 999

Text #7

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.7623. Min Act: -0.0000

Data Index: 1098118 (C4 (Web Text))

Max Activating Token Index: 418

Text #8

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.7510. Min Act: -0.0000

Data Index: 1162741 (C4 (Web Text))

Max Activating Token Index: 778

Text #9

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.6551. Min Act: -0.0000

Data Index: 15683 (C4 (Web Text))

Max Activating Token Index: 530

Text #10

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.8620. Min Act: -0.0000

Data Index: 197267 (C4 (Web Text))

Max Activating Token Index: 146

Text #11

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.7541. Min Act: -0.0000

Data Index: 1392512 (Python Code)

Max Activating Token Index: 564

Text #12

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.7321. Min Act: -0.0000

Data Index: 962574 (C4 (Web Text))

Max Activating Token Index: 673

Text #13

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.7237. Min Act: -0.0000

Data Index: 1464006 (Python Code)

Max Activating Token Index: 53

Text #14

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.6224. Min Act: -0.0000

Data Index: 362494 (C4 (Web Text))

Max Activating Token Index: 975

Text #15

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.6089. Min Act: -0.0000

Data Index: 173960 (C4 (Web Text))

Max Activating Token Index: 156

Text #16

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.6928. Min Act: -0.0000

Data Index: 1256233 (C4 (Web Text))

Max Activating Token Index: 612

Text #17

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.6874. Min Act: -0.0000

Data Index: 248218 (C4 (Web Text))

Max Activating Token Index: 977

Text #18

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.6899. Min Act: -0.0000

Data Index: 926645 (C4 (Web Text))

Max Activating Token Index: 803

Text #19

Max Range: 2.6932. Min Range: -2.6932

Max Act: 1.7639. Min Act: -0.0000

Data Index: 346033 (C4 (Web Text))

Max Activating Token Index: 805