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

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

Neuron 846 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: 0.0258. Min Range: -0.0258

Max Act: 0.0258. Min Act: -0.0006

Data Index: 631356 (C4 (Web Text))

Max Activating Token Index: 230

Text #1

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0245. Min Act: -0.0006

Data Index: 1596564 (Python Code)

Max Activating Token Index: 492

Text #2

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0229. Min Act: -0.0006

Data Index: 1372436 (Python Code)

Max Activating Token Index: 522

Text #3

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0242. Min Act: -0.0006

Data Index: 920292 (C4 (Web Text))

Max Activating Token Index: 879

Text #4

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0232. Min Act: -0.0006

Data Index: 149472 (C4 (Web Text))

Max Activating Token Index: 854

Text #5

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0226. Min Act: -0.0006

Data Index: 1529899 (Python Code)

Max Activating Token Index: 825

Text #6

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0229. Min Act: -0.0006

Data Index: 809766 (C4 (Web Text))

Max Activating Token Index: 336

Text #7

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0233. Min Act: -0.0006

Data Index: 1301096 (C4 (Web Text))

Max Activating Token Index: 766

Text #8

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0227. Min Act: -0.0005

Data Index: 1476231 (Python Code)

Max Activating Token Index: 428

Text #9

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0209. Min Act: -0.0006

Data Index: 1561852 (Python Code)

Max Activating Token Index: 601

Text #10

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0227. Min Act: -0.0006

Data Index: 914437 (C4 (Web Text))

Max Activating Token Index: 461

Text #11

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0221. Min Act: -0.0006

Data Index: 1015207 (C4 (Web Text))

Max Activating Token Index: 414

Text #12

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0221. Min Act: -0.0006

Data Index: 1495916 (Python Code)

Max Activating Token Index: 400

Text #13

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0214. Min Act: -0.0006

Data Index: 665962 (C4 (Web Text))

Max Activating Token Index: 834

Text #14

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0219. Min Act: -0.0005

Data Index: 1165496 (C4 (Web Text))

Max Activating Token Index: 201

Text #15

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0215. Min Act: -0.0005

Data Index: 311158 (C4 (Web Text))

Max Activating Token Index: 924

Text #16

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0215. Min Act: -0.0006

Data Index: 89626 (C4 (Web Text))

Max Activating Token Index: 946

Text #17

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0216. Min Act: -0.0006

Data Index: 1027973 (C4 (Web Text))

Max Activating Token Index: 651

Text #18

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0218. Min Act: -0.0006

Data Index: 16067 (C4 (Web Text))

Max Activating Token Index: 636

Text #19

Max Range: 0.0258. Min Range: -0.0258

Max Act: 0.0213. Min Act: -0.0006

Data Index: 1034628 (C4 (Web Text))

Max Activating Token Index: 832