Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 3393 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1578. Min Act: -0.0000

Data Index: 696097 (C4 (Web Text))

Max Activating Token Index: 772

Text #1

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1494. Min Act: -0.0000

Data Index: 483033 (C4 (Web Text))

Max Activating Token Index: 322

Text #2

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1474. Min Act: -0.0000

Data Index: 1305193 (C4 (Web Text))

Max Activating Token Index: 992

Text #3

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1423. Min Act: -0.0000

Data Index: 239670 (C4 (Web Text))

Max Activating Token Index: 645

Text #4

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1383. Min Act: -0.0000

Data Index: 1577261 (Python Code)

Max Activating Token Index: 431

Text #5

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1358. Min Act: -0.0000

Data Index: 129296 (C4 (Web Text))

Max Activating Token Index: 959

Text #6

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1295. Min Act: -0.0000

Data Index: 907877 (C4 (Web Text))

Max Activating Token Index: 123

Text #7

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1273. Min Act: -0.0000

Data Index: 403782 (C4 (Web Text))

Max Activating Token Index: 816

Text #8

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1182. Min Act: -0.0000

Data Index: 990480 (C4 (Web Text))

Max Activating Token Index: 48

Text #9

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1242. Min Act: -0.0000

Data Index: 822056 (C4 (Web Text))

Max Activating Token Index: 523

Text #10

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1322. Min Act: -0.0000

Data Index: 1156757 (C4 (Web Text))

Max Activating Token Index: 27

Text #11

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1165. Min Act: -0.0000

Data Index: 948 (C4 (Web Text))

Max Activating Token Index: 90

Text #12

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1267. Min Act: -0.0000

Data Index: 1005145 (C4 (Web Text))

Max Activating Token Index: 490

Text #13

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1118. Min Act: -0.0000

Data Index: 1084977 (C4 (Web Text))

Max Activating Token Index: 540

Text #14

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1251. Min Act: -0.0000

Data Index: 812819 (C4 (Web Text))

Max Activating Token Index: 787

Text #15

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1084. Min Act: -0.0000

Data Index: 1621848 (Python Code)

Max Activating Token Index: 179

Text #16

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1112. Min Act: -0.0000

Data Index: 281592 (C4 (Web Text))

Max Activating Token Index: 888

Text #17

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1187. Min Act: -0.0000

Data Index: 986629 (C4 (Web Text))

Max Activating Token Index: 345

Text #18

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1238. Min Act: -0.0000

Data Index: 371059 (C4 (Web Text))

Max Activating Token Index: 547

Text #19

Max Range: 0.1578. Min Range: -0.1578

Max Act: 0.1179. Min Act: -0.0000

Data Index: 1410298 (Python Code)

Max Activating Token Index: 308