Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 1092 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1473. Min Act: -0.0001

Data Index: 778345 (C4 (Web Text))

Max Activating Token Index: 873

Text #1

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1391. Min Act: -0.0001

Data Index: 919199 (C4 (Web Text))

Max Activating Token Index: 10

Text #2

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1336. Min Act: -0.0001

Data Index: 188681 (C4 (Web Text))

Max Activating Token Index: 582

Text #3

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1328. Min Act: -0.0002

Data Index: 738624 (C4 (Web Text))

Max Activating Token Index: 57

Text #4

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1302. Min Act: -0.0002

Data Index: 840760 (C4 (Web Text))

Max Activating Token Index: 139

Text #5

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1306. Min Act: -0.0002

Data Index: 160583 (C4 (Web Text))

Max Activating Token Index: 118

Text #6

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1277. Min Act: -0.0002

Data Index: 1185317 (C4 (Web Text))

Max Activating Token Index: 780

Text #7

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1213. Min Act: -0.0002

Data Index: 200748 (C4 (Web Text))

Max Activating Token Index: 608

Text #8

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1221. Min Act: -0.0001

Data Index: 881514 (C4 (Web Text))

Max Activating Token Index: 85

Text #9

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1221. Min Act: -0.0002

Data Index: 595303 (C4 (Web Text))

Max Activating Token Index: 909

Text #10

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1251. Min Act: -0.0002

Data Index: 111012 (C4 (Web Text))

Max Activating Token Index: 137

Text #11

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1252. Min Act: -0.0002

Data Index: 276094 (C4 (Web Text))

Max Activating Token Index: 798

Text #12

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1134. Min Act: -0.0002

Data Index: 824383 (C4 (Web Text))

Max Activating Token Index: 143

Text #13

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1207. Min Act: -0.0001

Data Index: 173378 (C4 (Web Text))

Max Activating Token Index: 113

Text #14

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1196. Min Act: -0.0001

Data Index: 1255373 (C4 (Web Text))

Max Activating Token Index: 591

Text #15

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1194. Min Act: -0.0001

Data Index: 575460 (C4 (Web Text))

Max Activating Token Index: 706

Text #16

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1185. Min Act: -0.0002

Data Index: 361241 (C4 (Web Text))

Max Activating Token Index: 473

Text #17

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1122. Min Act: -0.0002

Data Index: 151885 (C4 (Web Text))

Max Activating Token Index: 274

Text #18

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1175. Min Act: -0.0002

Data Index: 505146 (C4 (Web Text))

Max Activating Token Index: 40

Text #19

Max Range: 0.1473. Min Range: -0.1473

Max Act: 0.1169. Min Act: -0.0002

Data Index: 1293961 (C4 (Web Text))

Max Activating Token Index: 47