Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 5000 in Layer 10

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 4.0052. Min Range: -4.0052

Max Act: 4.0052. Min Act: -0.0000

Data Index: 49712 (C4 (Web Text))

Max Activating Token Index: 96

Text #1

Max Range: 4.0052. Min Range: -4.0052

Max Act: 3.5695. Min Act: -0.0000

Data Index: 725890 (C4 (Web Text))

Max Activating Token Index: 1016

Text #2

Max Range: 4.0052. Min Range: -4.0052

Max Act: 3.2030. Min Act: -0.0000

Data Index: 269765 (C4 (Web Text))

Max Activating Token Index: 510

Text #3

Max Range: 4.0052. Min Range: -4.0052

Max Act: 3.1827. Min Act: -0.0000

Data Index: 216109 (C4 (Web Text))

Max Activating Token Index: 584

Text #4

Max Range: 4.0052. Min Range: -4.0052

Max Act: 3.0478. Min Act: -0.0000

Data Index: 339118 (C4 (Web Text))

Max Activating Token Index: 741

Text #5

Max Range: 4.0052. Min Range: -4.0052

Max Act: 2.8382. Min Act: -0.0000

Data Index: 825154 (C4 (Web Text))

Max Activating Token Index: 313

Text #6

Max Range: 4.0052. Min Range: -4.0052

Max Act: 3.0816. Min Act: -0.0000

Data Index: 1174427 (C4 (Web Text))

Max Activating Token Index: 975

Text #7

Max Range: 4.0052. Min Range: -4.0052

Max Act: 2.8687. Min Act: -0.0000

Data Index: 1221883 (C4 (Web Text))

Max Activating Token Index: 627

Text #8

Max Range: 4.0052. Min Range: -4.0052

Max Act: 3.0021. Min Act: -0.0000

Data Index: 401109 (C4 (Web Text))

Max Activating Token Index: 123

Text #9

Max Range: 4.0052. Min Range: -4.0052

Max Act: 2.9825. Min Act: -0.0000

Data Index: 377447 (C4 (Web Text))

Max Activating Token Index: 455

Text #10

Max Range: 4.0052. Min Range: -4.0052

Max Act: 3.0007. Min Act: -0.0000

Data Index: 36058 (C4 (Web Text))

Max Activating Token Index: 919

Text #11

Max Range: 4.0052. Min Range: -4.0052

Max Act: 2.8304. Min Act: -0.0000

Data Index: 169916 (C4 (Web Text))

Max Activating Token Index: 64

Text #12

Max Range: 4.0052. Min Range: -4.0052

Max Act: 2.9430. Min Act: -0.0000

Data Index: 294599 (C4 (Web Text))

Max Activating Token Index: 380

Text #13

Max Range: 4.0052. Min Range: -4.0052

Max Act: 2.9285. Min Act: -0.0000

Data Index: 599309 (C4 (Web Text))

Max Activating Token Index: 294

Text #14

Max Range: 4.0052. Min Range: -4.0052

Max Act: 2.9234. Min Act: -0.0000

Data Index: 62556 (C4 (Web Text))

Max Activating Token Index: 230

Text #15

Max Range: 4.0052. Min Range: -4.0052

Max Act: 2.8971. Min Act: -0.0000

Data Index: 381174 (C4 (Web Text))

Max Activating Token Index: 600

Text #16

Max Range: 4.0052. Min Range: -4.0052

Max Act: 3.0354. Min Act: -0.0000

Data Index: 1195854 (C4 (Web Text))

Max Activating Token Index: 172

Text #17

Max Range: 4.0052. Min Range: -4.0052

Max Act: 2.8824. Min Act: -0.0000

Data Index: 629073 (C4 (Web Text))

Max Activating Token Index: 161

Text #18

Max Range: 4.0052. Min Range: -4.0052

Max Act: 2.8981. Min Act: -0.0000

Data Index: 828993 (C4 (Web Text))

Max Activating Token Index: 158

Text #19

Max Range: 4.0052. Min Range: -4.0052

Max Act: 2.8870. Min Act: -0.0000

Data Index: 1229614 (C4 (Web Text))

Max Activating Token Index: 812