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

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

Neuron 3755 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: 0.9168. Min Range: -0.9168

Max Act: 0.9168. Min Act: -0.0001

Data Index: 905371 (C4 (Web Text))

Max Activating Token Index: 860

Text #1

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.8591. Min Act: -0.0001

Data Index: 357009 (C4 (Web Text))

Max Activating Token Index: 694

Text #2

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.8225. Min Act: -0.0001

Data Index: 1044029 (C4 (Web Text))

Max Activating Token Index: 519

Text #3

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7998. Min Act: -0.0001

Data Index: 883086 (C4 (Web Text))

Max Activating Token Index: 770

Text #4

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7447. Min Act: -0.0001

Data Index: 1031119 (C4 (Web Text))

Max Activating Token Index: 26

Text #5

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7457. Min Act: -0.0001

Data Index: 91998 (C4 (Web Text))

Max Activating Token Index: 369

Text #6

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7369. Min Act: -0.0001

Data Index: 462405 (C4 (Web Text))

Max Activating Token Index: 30

Text #7

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7658. Min Act: -0.0001

Data Index: 198770 (C4 (Web Text))

Max Activating Token Index: 880

Text #8

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7623. Min Act: -0.0001

Data Index: 59462 (C4 (Web Text))

Max Activating Token Index: 20

Text #9

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7630. Min Act: -0.0001

Data Index: 455776 (C4 (Web Text))

Max Activating Token Index: 20

Text #10

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7607. Min Act: -0.0001

Data Index: 208081 (C4 (Web Text))

Max Activating Token Index: 921

Text #11

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7225. Min Act: -0.0001

Data Index: 167785 (C4 (Web Text))

Max Activating Token Index: 631

Text #12

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7435. Min Act: -0.0001

Data Index: 1031022 (C4 (Web Text))

Max Activating Token Index: 281

Text #13

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.6874. Min Act: -0.0001

Data Index: 1125821 (C4 (Web Text))

Max Activating Token Index: 657

Text #14

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7004. Min Act: -0.0001

Data Index: 1333214 (C4 (Web Text))

Max Activating Token Index: 680

Text #15

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.6317. Min Act: -0.0001

Data Index: 1517061 (Python Code)

Max Activating Token Index: 59

Text #16

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7468. Min Act: -0.0001

Data Index: 400859 (C4 (Web Text))

Max Activating Token Index: 714

Text #17

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.6958. Min Act: -0.0001

Data Index: 1218272 (C4 (Web Text))

Max Activating Token Index: 259

Text #18

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.7563. Min Act: -0.0001

Data Index: 1088491 (C4 (Web Text))

Max Activating Token Index: 715

Text #19

Max Range: 0.9168. Min Range: -0.9168

Max Act: 0.6732. Min Act: -0.0001

Data Index: 614050 (C4 (Web Text))

Max Activating Token Index: 653