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

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

Neuron 3971 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: 0.4267. Min Range: -0.4267

Max Act: 0.1691. Min Act: -0.0000

Data Index: 1400650 (Python Code)

Max Activating Token Index: 188

Text #1

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.4267. Min Act: -0.0000

Data Index: 1083842 (C4 (Web Text))

Max Activating Token Index: 648

Text #2

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1465. Min Act: -0.0000

Data Index: 1600082 (Python Code)

Max Activating Token Index: 888

Text #3

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1279. Min Act: -0.0000

Data Index: 793066 (C4 (Web Text))

Max Activating Token Index: 623

Text #4

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1292. Min Act: -0.0000

Data Index: 1370327 (Python Code)

Max Activating Token Index: 283

Text #5

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1441. Min Act: -0.0000

Data Index: 1462619 (Python Code)

Max Activating Token Index: 71

Text #6

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1241. Min Act: -0.0000

Data Index: 1305879 (C4 (Web Text))

Max Activating Token Index: 907

Text #7

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1322. Min Act: -0.0000

Data Index: 510233 (C4 (Web Text))

Max Activating Token Index: 356

Text #8

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1103. Min Act: -0.0000

Data Index: 1449100 (Python Code)

Max Activating Token Index: 209

Text #9

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1266. Min Act: -0.0000

Data Index: 679800 (C4 (Web Text))

Max Activating Token Index: 508

Text #10

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1231. Min Act: -0.0000

Data Index: 922412 (C4 (Web Text))

Max Activating Token Index: 142

Text #11

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1355. Min Act: -0.0000

Data Index: 1433970 (Python Code)

Max Activating Token Index: 328

Text #12

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1204. Min Act: -0.0000

Data Index: 1647651 (Python Code)

Max Activating Token Index: 327

Text #13

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1204. Min Act: -0.0000

Data Index: 1485608 (Python Code)

Max Activating Token Index: 984

Text #14

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1112. Min Act: -0.0000

Data Index: 1107337 (C4 (Web Text))

Max Activating Token Index: 822

Text #15

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1232. Min Act: -0.0000

Data Index: 1305607 (C4 (Web Text))

Max Activating Token Index: 852

Text #16

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1150. Min Act: -0.0000

Data Index: 625135 (C4 (Web Text))

Max Activating Token Index: 791

Text #17

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1186. Min Act: -0.0000

Data Index: 840765 (C4 (Web Text))

Max Activating Token Index: 130

Text #18

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1133. Min Act: -0.0000

Data Index: 338903 (C4 (Web Text))

Max Activating Token Index: 828

Text #19

Max Range: 0.4267. Min Range: -0.4267

Max Act: 0.1168. Min Act: -0.0000

Data Index: 119156 (C4 (Web Text))

Max Activating Token Index: 657