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

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

Neuron 2000 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: 0.0220. Min Range: -0.0220

Max Act: 0.0220. Min Act: -0.0006

Data Index: 830637 (C4 (Web Text))

Max Activating Token Index: 535

Text #1

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0213. Min Act: -0.0006

Data Index: 3300 (C4 (Web Text))

Max Activating Token Index: 603

Text #2

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0211. Min Act: -0.0006

Data Index: 653485 (C4 (Web Text))

Max Activating Token Index: 464

Text #3

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0200. Min Act: -0.0006

Data Index: 507136 (C4 (Web Text))

Max Activating Token Index: 20

Text #4

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0199. Min Act: -0.0006

Data Index: 796161 (C4 (Web Text))

Max Activating Token Index: 612

Text #5

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0197. Min Act: -0.0006

Data Index: 1172842 (C4 (Web Text))

Max Activating Token Index: 310

Text #6

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0198. Min Act: -0.0006

Data Index: 356241 (C4 (Web Text))

Max Activating Token Index: 846

Text #7

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0192. Min Act: -0.0006

Data Index: 964599 (C4 (Web Text))

Max Activating Token Index: 461

Text #8

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0199. Min Act: -0.0006

Data Index: 710531 (C4 (Web Text))

Max Activating Token Index: 810

Text #9

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0191. Min Act: -0.0006

Data Index: 685028 (C4 (Web Text))

Max Activating Token Index: 670

Text #10

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0191. Min Act: -0.0006

Data Index: 46237 (C4 (Web Text))

Max Activating Token Index: 336

Text #11

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0193. Min Act: -0.0006

Data Index: 637904 (C4 (Web Text))

Max Activating Token Index: 86

Text #12

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0193. Min Act: -0.0006

Data Index: 426297 (C4 (Web Text))

Max Activating Token Index: 761

Text #13

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0189. Min Act: -0.0006

Data Index: 335107 (C4 (Web Text))

Max Activating Token Index: 829

Text #14

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0192. Min Act: -0.0006

Data Index: 942475 (C4 (Web Text))

Max Activating Token Index: 604

Text #15

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0191. Min Act: -0.0006

Data Index: 1228640 (C4 (Web Text))

Max Activating Token Index: 105

Text #16

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0195. Min Act: -0.0006

Data Index: 901830 (C4 (Web Text))

Max Activating Token Index: 579

Text #17

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0186. Min Act: -0.0006

Data Index: 556905 (C4 (Web Text))

Max Activating Token Index: 447

Text #18

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0188. Min Act: -0.0006

Data Index: 120288 (C4 (Web Text))

Max Activating Token Index: 98

Text #19

Max Range: 0.0220. Min Range: -0.0220

Max Act: 0.0189. Min Act: -0.0006

Data Index: 1267956 (C4 (Web Text))

Max Activating Token Index: 997