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

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

Neuron 490 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: 0.0208. Min Range: -0.0208

Max Act: 0.0208. Min Act: -0.0006

Data Index: 1148000 (C4 (Web Text))

Max Activating Token Index: 1023

Text #1

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0191. Min Act: -0.0006

Data Index: 1147784 (C4 (Web Text))

Max Activating Token Index: 1023

Text #2

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0187. Min Act: -0.0006

Data Index: 264003 (C4 (Web Text))

Max Activating Token Index: 1023

Text #3

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0185. Min Act: -0.0006

Data Index: 1388719 (Python Code)

Max Activating Token Index: 692

Text #4

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0180. Min Act: -0.0006

Data Index: 1447416 (Python Code)

Max Activating Token Index: 114

Text #5

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0175. Min Act: -0.0006

Data Index: 1073101 (C4 (Web Text))

Max Activating Token Index: 1023

Text #6

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0172. Min Act: -0.0006

Data Index: 1210603 (C4 (Web Text))

Max Activating Token Index: 1023

Text #7

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0165. Min Act: -0.0006

Data Index: 12804 (C4 (Web Text))

Max Activating Token Index: 1023

Text #8

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0161. Min Act: -0.0006

Data Index: 335922 (C4 (Web Text))

Max Activating Token Index: 1023

Text #9

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0153. Min Act: -0.0006

Data Index: 128824 (C4 (Web Text))

Max Activating Token Index: 1023

Text #10

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0156. Min Act: -0.0006

Data Index: 1574747 (Python Code)

Max Activating Token Index: 145

Text #11

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0156. Min Act: -0.0006

Data Index: 843840 (C4 (Web Text))

Max Activating Token Index: 1023

Text #12

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0153. Min Act: -0.0006

Data Index: 117695 (C4 (Web Text))

Max Activating Token Index: 1023

Text #13

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0151. Min Act: -0.0006

Data Index: 448130 (C4 (Web Text))

Max Activating Token Index: 1023

Text #14

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0153. Min Act: -0.0006

Data Index: 1465276 (Python Code)

Max Activating Token Index: 141

Text #15

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0152. Min Act: -0.0006

Data Index: 27803 (C4 (Web Text))

Max Activating Token Index: 1023

Text #16

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0147. Min Act: -0.0006

Data Index: 575104 (C4 (Web Text))

Max Activating Token Index: 1023

Text #17

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0150. Min Act: -0.0006

Data Index: 773332 (C4 (Web Text))

Max Activating Token Index: 1023

Text #18

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0143. Min Act: -0.0006

Data Index: 1313310 (C4 (Web Text))

Max Activating Token Index: 1023

Text #19

Max Range: 0.0208. Min Range: -0.0208

Max Act: 0.0146. Min Act: -0.0006

Data Index: 920503 (C4 (Web Text))

Max Activating Token Index: 1023