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

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

Neuron 234 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0200. Min Act: -0.0006

Data Index: 230017 (C4 (Web Text))

Max Activating Token Index: 298

Text #1

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0176. Min Act: -0.0006

Data Index: 77572 (C4 (Web Text))

Max Activating Token Index: 563

Text #2

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0177. Min Act: -0.0006

Data Index: 477872 (C4 (Web Text))

Max Activating Token Index: 212

Text #3

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0169. Min Act: -0.0006

Data Index: 332347 (C4 (Web Text))

Max Activating Token Index: 29

Text #4

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0173. Min Act: -0.0006

Data Index: 1133460 (C4 (Web Text))

Max Activating Token Index: 735

Text #5

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0168. Min Act: -0.0006

Data Index: 1264063 (C4 (Web Text))

Max Activating Token Index: 321

Text #6

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0166. Min Act: -0.0006

Data Index: 621438 (C4 (Web Text))

Max Activating Token Index: 603

Text #7

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0159. Min Act: -0.0006

Data Index: 83970 (C4 (Web Text))

Max Activating Token Index: 373

Text #8

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0157. Min Act: -0.0006

Data Index: 888824 (C4 (Web Text))

Max Activating Token Index: 96

Text #9

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0153. Min Act: -0.0006

Data Index: 1128947 (C4 (Web Text))

Max Activating Token Index: 717

Text #10

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0155. Min Act: -0.0006

Data Index: 928210 (C4 (Web Text))

Max Activating Token Index: 157

Text #11

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0161. Min Act: -0.0006

Data Index: 605511 (C4 (Web Text))

Max Activating Token Index: 1009

Text #12

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0154. Min Act: -0.0006

Data Index: 859375 (C4 (Web Text))

Max Activating Token Index: 57

Text #13

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0152. Min Act: -0.0006

Data Index: 713662 (C4 (Web Text))

Max Activating Token Index: 850

Text #14

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0156. Min Act: -0.0006

Data Index: 1604093 (Python Code)

Max Activating Token Index: 481

Text #15

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0154. Min Act: -0.0006

Data Index: 588734 (C4 (Web Text))

Max Activating Token Index: 986

Text #16

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0151. Min Act: -0.0006

Data Index: 837365 (C4 (Web Text))

Max Activating Token Index: 876

Text #17

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0151. Min Act: -0.0006

Data Index: 112242 (C4 (Web Text))

Max Activating Token Index: 803

Text #18

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0153. Min Act: -0.0006

Data Index: 1219866 (C4 (Web Text))

Max Activating Token Index: 83

Text #19

Max Range: 0.0200. Min Range: -0.0200

Max Act: 0.0144. Min Act: -0.0006

Data Index: 232164 (C4 (Web Text))

Max Activating Token Index: 807