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

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

Neuron 4608 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0164. Min Act: -0.0001

Data Index: 976721 (C4 (Web Text))

Max Activating Token Index: 886

Text #1

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0187. Min Act: -0.0001

Data Index: 510782 (C4 (Web Text))

Max Activating Token Index: 12

Text #2

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0178. Min Act: -0.0001

Data Index: 200127 (C4 (Web Text))

Max Activating Token Index: 659

Text #3

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0174. Min Act: -0.0001

Data Index: 180248 (C4 (Web Text))

Max Activating Token Index: 266

Text #4

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0167. Min Act: -0.0001

Data Index: 552702 (C4 (Web Text))

Max Activating Token Index: 912

Text #5

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0164. Min Act: -0.0001

Data Index: 1079155 (C4 (Web Text))

Max Activating Token Index: 7

Text #6

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0171. Min Act: -0.0001

Data Index: 749000 (C4 (Web Text))

Max Activating Token Index: 987

Text #7

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0170. Min Act: -0.0001

Data Index: 1341045 (C4 (Web Text))

Max Activating Token Index: 348

Text #8

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0175. Min Act: -0.0001

Data Index: 1013334 (C4 (Web Text))

Max Activating Token Index: 35

Text #9

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0132. Min Act: -0.0001

Data Index: 1616971 (Python Code)

Max Activating Token Index: 678

Text #10

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0150. Min Act: -0.0001

Data Index: 1238750 (C4 (Web Text))

Max Activating Token Index: 692

Text #11

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0160. Min Act: -0.0001

Data Index: 455455 (C4 (Web Text))

Max Activating Token Index: 314

Text #12

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0153. Min Act: -0.0001

Data Index: 528191 (C4 (Web Text))

Max Activating Token Index: 226

Text #13

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0155. Min Act: -0.0001

Data Index: 3910 (C4 (Web Text))

Max Activating Token Index: 73

Text #14

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0147. Min Act: -0.0001

Data Index: 44366 (C4 (Web Text))

Max Activating Token Index: 45

Text #15

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0151. Min Act: -0.0001

Data Index: 1367087 (Python Code)

Max Activating Token Index: 198

Text #16

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0150. Min Act: -0.0001

Data Index: 870803 (C4 (Web Text))

Max Activating Token Index: 128

Text #17

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0153. Min Act: -0.0001

Data Index: 958433 (C4 (Web Text))

Max Activating Token Index: 178

Text #18

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0142. Min Act: -0.0001

Data Index: 228355 (C4 (Web Text))

Max Activating Token Index: 72

Text #19

Max Range: 0.0187. Min Range: -0.0187

Max Act: 0.0144. Min Act: -0.0001

Data Index: 1630201 (Python Code)

Max Activating Token Index: 700