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

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

Neuron 20 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.4096. Min Range: -3.4096

Max Act: 3.4096. Min Act: -0.0000

Data Index: 434257 (C4 (Web Text))

Max Activating Token Index: 794

Text #1

Max Range: 3.4096. Min Range: -3.4096

Max Act: 3.0722. Min Act: -0.0000

Data Index: 232803 (C4 (Web Text))

Max Activating Token Index: 739

Text #2

Max Range: 3.4096. Min Range: -3.4096

Max Act: 3.0414. Min Act: -0.0000

Data Index: 496165 (C4 (Web Text))

Max Activating Token Index: 572

Text #3

Max Range: 3.4096. Min Range: -3.4096

Max Act: 3.1368. Min Act: -0.0000

Data Index: 515135 (C4 (Web Text))

Max Activating Token Index: 1008

Text #4

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.9375. Min Act: -0.0000

Data Index: 1164628 (C4 (Web Text))

Max Activating Token Index: 159

Text #5

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.8052. Min Act: -0.0000

Data Index: 270259 (C4 (Web Text))

Max Activating Token Index: 231

Text #6

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.7974. Min Act: -0.0000

Data Index: 685339 (C4 (Web Text))

Max Activating Token Index: 566

Text #7

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.6465. Min Act: -0.0000

Data Index: 41783 (C4 (Web Text))

Max Activating Token Index: 158

Text #8

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.7190. Min Act: -0.0000

Data Index: 655380 (C4 (Web Text))

Max Activating Token Index: 544

Text #9

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.6713. Min Act: -0.0000

Data Index: 512916 (C4 (Web Text))

Max Activating Token Index: 51

Text #10

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.6499. Min Act: -0.0000

Data Index: 449603 (C4 (Web Text))

Max Activating Token Index: 827

Text #11

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.4425. Min Act: -0.0000

Data Index: 952163 (C4 (Web Text))

Max Activating Token Index: 693

Text #12

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.6606. Min Act: -0.0000

Data Index: 255687 (C4 (Web Text))

Max Activating Token Index: 196

Text #13

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.3529. Min Act: -0.0000

Data Index: 425924 (C4 (Web Text))

Max Activating Token Index: 249

Text #14

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.5694. Min Act: -0.0000

Data Index: 588327 (C4 (Web Text))

Max Activating Token Index: 224

Text #15

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.2866. Min Act: -0.0000

Data Index: 612137 (C4 (Web Text))

Max Activating Token Index: 303

Text #16

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.4179. Min Act: -0.0000

Data Index: 1597877 (Python Code)

Max Activating Token Index: 127

Text #17

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.4218. Min Act: -0.0000

Data Index: 24315 (C4 (Web Text))

Max Activating Token Index: 411

Text #18

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.3608. Min Act: -0.0000

Data Index: 252426 (C4 (Web Text))

Max Activating Token Index: 193

Text #19

Max Range: 3.4096. Min Range: -3.4096

Max Act: 2.3486. Min Act: -0.0000

Data Index: 1597847 (Python Code)

Max Activating Token Index: 391