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

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

Neuron 757 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0927. Min Act: -0.0000

Data Index: 97690 (C4 (Web Text))

Max Activating Token Index: 385

Text #1

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0425. Min Act: -0.0000

Data Index: 1247510 (C4 (Web Text))

Max Activating Token Index: 37

Text #2

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0245. Min Act: -0.0000

Data Index: 112284 (C4 (Web Text))

Max Activating Token Index: 278

Text #3

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0259. Min Act: -0.0000

Data Index: 241401 (C4 (Web Text))

Max Activating Token Index: 497

Text #4

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0255. Min Act: -0.0000

Data Index: 108167 (C4 (Web Text))

Max Activating Token Index: 240

Text #5

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0210. Min Act: -0.0000

Data Index: 669637 (C4 (Web Text))

Max Activating Token Index: 341

Text #6

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0418. Min Act: -0.0000

Data Index: 598960 (C4 (Web Text))

Max Activating Token Index: 782

Text #7

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0204. Min Act: -0.0000

Data Index: 636690 (C4 (Web Text))

Max Activating Token Index: 275

Text #8

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0159. Min Act: -0.0000

Data Index: 1129842 (C4 (Web Text))

Max Activating Token Index: 165

Text #9

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0201. Min Act: -0.0000

Data Index: 1140960 (C4 (Web Text))

Max Activating Token Index: 813

Text #10

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0167. Min Act: -0.0000

Data Index: 946315 (C4 (Web Text))

Max Activating Token Index: 772

Text #11

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0181. Min Act: -0.0000

Data Index: 1108890 (C4 (Web Text))

Max Activating Token Index: 298

Text #12

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0176. Min Act: -0.0000

Data Index: 1369282 (Python Code)

Max Activating Token Index: 835

Text #13

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0170. Min Act: -0.0000

Data Index: 764299 (C4 (Web Text))

Max Activating Token Index: 451

Text #14

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0172. Min Act: -0.0000

Data Index: 934141 (C4 (Web Text))

Max Activating Token Index: 278

Text #15

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0175. Min Act: -0.0000

Data Index: 62874 (C4 (Web Text))

Max Activating Token Index: 441

Text #16

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0167. Min Act: -0.0000

Data Index: 1551207 (Python Code)

Max Activating Token Index: 788

Text #17

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0171. Min Act: -0.0000

Data Index: 234156 (C4 (Web Text))

Max Activating Token Index: 1003

Text #18

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0166. Min Act: -0.0000

Data Index: 1049457 (C4 (Web Text))

Max Activating Token Index: 310

Text #19

Max Range: 0.0927. Min Range: -0.0927

Max Act: 0.0168. Min Act: -0.0000

Data Index: 912077 (C4 (Web Text))

Max Activating Token Index: 899