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

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

Neuron 480 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: 0.0222. Min Range: -0.0222

Max Act: 0.0222. Min Act: -0.0006

Data Index: 1380780 (Python Code)

Max Activating Token Index: 181

Text #1

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0191. Min Act: -0.0006

Data Index: 1508025 (Python Code)

Max Activating Token Index: 186

Text #2

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0170. Min Act: -0.0006

Data Index: 1384750 (Python Code)

Max Activating Token Index: 429

Text #3

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0180. Min Act: -0.0006

Data Index: 155575 (C4 (Web Text))

Max Activating Token Index: 911

Text #4

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0183. Min Act: -0.0006

Data Index: 1587652 (Python Code)

Max Activating Token Index: 1018

Text #5

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0174. Min Act: -0.0006

Data Index: 1403283 (Python Code)

Max Activating Token Index: 758

Text #6

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0188. Min Act: -0.0006

Data Index: 932722 (C4 (Web Text))

Max Activating Token Index: 120

Text #7

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0167. Min Act: -0.0006

Data Index: 1324550 (C4 (Web Text))

Max Activating Token Index: 815

Text #8

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0159. Min Act: -0.0006

Data Index: 1536301 (Python Code)

Max Activating Token Index: 570

Text #9

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0170. Min Act: -0.0006

Data Index: 524220 (C4 (Web Text))

Max Activating Token Index: 431

Text #10

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0165. Min Act: -0.0006

Data Index: 1259509 (C4 (Web Text))

Max Activating Token Index: 27

Text #11

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0168. Min Act: -0.0006

Data Index: 1530545 (Python Code)

Max Activating Token Index: 864

Text #12

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0161. Min Act: -0.0006

Data Index: 745264 (C4 (Web Text))

Max Activating Token Index: 995

Text #13

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0164. Min Act: -0.0006

Data Index: 1025866 (C4 (Web Text))

Max Activating Token Index: 344

Text #14

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0153. Min Act: -0.0006

Data Index: 1373791 (Python Code)

Max Activating Token Index: 436

Text #15

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0171. Min Act: -0.0006

Data Index: 3191 (C4 (Web Text))

Max Activating Token Index: 885

Text #16

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0158. Min Act: -0.0006

Data Index: 1406379 (Python Code)

Max Activating Token Index: 39

Text #17

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0142. Min Act: -0.0006

Data Index: 1413738 (Python Code)

Max Activating Token Index: 745

Text #18

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0156. Min Act: -0.0006

Data Index: 500658 (C4 (Web Text))

Max Activating Token Index: 971

Text #19

Max Range: 0.0222. Min Range: -0.0222

Max Act: 0.0152. Min Act: -0.0006

Data Index: 1081888 (C4 (Web Text))

Max Activating Token Index: 363