Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 2024 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0231. Min Act: -0.0006

Data Index: 215 (C4 (Web Text))

Max Activating Token Index: 681

Text #1

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0211. Min Act: -0.0006

Data Index: 770989 (C4 (Web Text))

Max Activating Token Index: 453

Text #2

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0214. Min Act: -0.0006

Data Index: 738410 (C4 (Web Text))

Max Activating Token Index: 905

Text #3

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0202. Min Act: -0.0006

Data Index: 719057 (C4 (Web Text))

Max Activating Token Index: 977

Text #4

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0207. Min Act: -0.0006

Data Index: 1051350 (C4 (Web Text))

Max Activating Token Index: 734

Text #5

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0203. Min Act: -0.0006

Data Index: 544684 (C4 (Web Text))

Max Activating Token Index: 908

Text #6

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0192. Min Act: -0.0006

Data Index: 836702 (C4 (Web Text))

Max Activating Token Index: 1015

Text #7

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0187. Min Act: -0.0006

Data Index: 763155 (C4 (Web Text))

Max Activating Token Index: 949

Text #8

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0187. Min Act: -0.0006

Data Index: 339530 (C4 (Web Text))

Max Activating Token Index: 55

Text #9

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0184. Min Act: -0.0006

Data Index: 925977 (C4 (Web Text))

Max Activating Token Index: 502

Text #10

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0184. Min Act: -0.0006

Data Index: 29340 (C4 (Web Text))

Max Activating Token Index: 914

Text #11

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0179. Min Act: -0.0006

Data Index: 411862 (C4 (Web Text))

Max Activating Token Index: 568

Text #12

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0172. Min Act: -0.0006

Data Index: 1194198 (C4 (Web Text))

Max Activating Token Index: 288

Text #13

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0166. Min Act: -0.0006

Data Index: 1183822 (C4 (Web Text))

Max Activating Token Index: 104

Text #14

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0172. Min Act: -0.0006

Data Index: 434414 (C4 (Web Text))

Max Activating Token Index: 495

Text #15

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0168. Min Act: -0.0006

Data Index: 448773 (C4 (Web Text))

Max Activating Token Index: 477

Text #16

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0167. Min Act: -0.0006

Data Index: 531594 (C4 (Web Text))

Max Activating Token Index: 468

Text #17

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0164. Min Act: -0.0006

Data Index: 1121861 (C4 (Web Text))

Max Activating Token Index: 153

Text #18

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0163. Min Act: -0.0006

Data Index: 1337560 (C4 (Web Text))

Max Activating Token Index: 89

Text #19

Max Range: 0.0231. Min Range: -0.0231

Max Act: 0.0166. Min Act: -0.0006

Data Index: 1035041 (C4 (Web Text))

Max Activating Token Index: 963