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

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

Neuron 12 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: 0.0267. Min Range: -0.0267

Max Act: 0.0266. Min Act: -0.0000

Data Index: 1381442 (Python Code)

Max Activating Token Index: 458

Text #1

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0267. Min Act: -0.0000

Data Index: 726222 (C4 (Web Text))

Max Activating Token Index: 315

Text #2

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0246. Min Act: -0.0000

Data Index: 1022557 (C4 (Web Text))

Max Activating Token Index: 146

Text #3

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0237. Min Act: -0.0000

Data Index: 368586 (C4 (Web Text))

Max Activating Token Index: 888

Text #4

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0227. Min Act: -0.0000

Data Index: 446621 (C4 (Web Text))

Max Activating Token Index: 966

Text #5

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0212. Min Act: -0.0000

Data Index: 629854 (C4 (Web Text))

Max Activating Token Index: 88

Text #6

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0226. Min Act: -0.0000

Data Index: 693724 (C4 (Web Text))

Max Activating Token Index: 429

Text #7

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0210. Min Act: -0.0000

Data Index: 714686 (C4 (Web Text))

Max Activating Token Index: 297

Text #8

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0212. Min Act: -0.0000

Data Index: 588029 (C4 (Web Text))

Max Activating Token Index: 268

Text #9

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0202. Min Act: -0.0000

Data Index: 433584 (C4 (Web Text))

Max Activating Token Index: 339

Text #10

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0211. Min Act: -0.0000

Data Index: 167107 (C4 (Web Text))

Max Activating Token Index: 195

Text #11

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0200. Min Act: -0.0000

Data Index: 1054102 (C4 (Web Text))

Max Activating Token Index: 480

Text #12

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0228. Min Act: -0.0000

Data Index: 22165 (C4 (Web Text))

Max Activating Token Index: 294

Text #13

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0212. Min Act: -0.0000

Data Index: 28574 (C4 (Web Text))

Max Activating Token Index: 729

Text #14

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0201. Min Act: -0.0000

Data Index: 1384134 (Python Code)

Max Activating Token Index: 820

Text #15

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0181. Min Act: -0.0000

Data Index: 482925 (C4 (Web Text))

Max Activating Token Index: 374

Text #16

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0197. Min Act: -0.0000

Data Index: 696114 (C4 (Web Text))

Max Activating Token Index: 623

Text #17

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0197. Min Act: -0.0000

Data Index: 1018203 (C4 (Web Text))

Max Activating Token Index: 394

Text #18

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0201. Min Act: -0.0000

Data Index: 62982 (C4 (Web Text))

Max Activating Token Index: 340

Text #19

Max Range: 0.0267. Min Range: -0.0267

Max Act: 0.0194. Min Act: -0.0000

Data Index: 452927 (C4 (Web Text))

Max Activating Token Index: 378