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

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

Neuron 1316 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.1321. Min Act: -0.0000

Data Index: 824168 (C4 (Web Text))

Max Activating Token Index: 582

Text #1

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.1287. Min Act: -0.0000

Data Index: 480168 (C4 (Web Text))

Max Activating Token Index: 796

Text #2

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.1161. Min Act: -0.0001

Data Index: 791576 (C4 (Web Text))

Max Activating Token Index: 778

Text #3

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.1225. Min Act: -0.0001

Data Index: 729151 (C4 (Web Text))

Max Activating Token Index: 980

Text #4

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.1216. Min Act: -0.0000

Data Index: 1138701 (C4 (Web Text))

Max Activating Token Index: 818

Text #5

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.1156. Min Act: -0.0000

Data Index: 355708 (C4 (Web Text))

Max Activating Token Index: 889

Text #6

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.0930. Min Act: -0.0001

Data Index: 1484627 (Python Code)

Max Activating Token Index: 715

Text #7

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.1097. Min Act: -0.0000

Data Index: 938019 (C4 (Web Text))

Max Activating Token Index: 314

Text #8

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.1104. Min Act: -0.0001

Data Index: 1543208 (Python Code)

Max Activating Token Index: 636

Text #9

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.0958. Min Act: -0.0001

Data Index: 1579468 (Python Code)

Max Activating Token Index: 910

Text #10

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.1030. Min Act: -0.0000

Data Index: 1007037 (C4 (Web Text))

Max Activating Token Index: 298

Text #11

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.0978. Min Act: -0.0001

Data Index: 1652721 (Python Code)

Max Activating Token Index: 532

Text #12

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.1002. Min Act: -0.0000

Data Index: 785958 (C4 (Web Text))

Max Activating Token Index: 967

Text #13

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.0971. Min Act: -0.0001

Data Index: 1614571 (Python Code)

Max Activating Token Index: 963

Text #14

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.1062. Min Act: -0.0001

Data Index: 1438920 (Python Code)

Max Activating Token Index: 688

Text #15

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.1032. Min Act: -0.0001

Data Index: 7634 (C4 (Web Text))

Max Activating Token Index: 104

Text #16

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.0921. Min Act: -0.0000

Data Index: 1236682 (C4 (Web Text))

Max Activating Token Index: 255

Text #17

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.0978. Min Act: -0.0000

Data Index: 592274 (C4 (Web Text))

Max Activating Token Index: 657

Text #18

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.0977. Min Act: -0.0001

Data Index: 1648161 (Python Code)

Max Activating Token Index: 949

Text #19

Max Range: 0.1321. Min Range: -0.1321

Max Act: 0.0903. Min Act: -0.0001

Data Index: 1656401 (Python Code)

Max Activating Token Index: 676