Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 120 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 3.4627. Min Range: -3.4627

Max Act: 3.2986. Min Act: -0.0000

Data Index: 552992 (C4 (Web Text))

Max Activating Token Index: 40

Text #1

Max Range: 3.4627. Min Range: -3.4627

Max Act: 3.4627. Min Act: -0.0000

Data Index: 874569 (C4 (Web Text))

Max Activating Token Index: 530

Text #2

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.5790. Min Act: -0.0000

Data Index: 841053 (C4 (Web Text))

Max Activating Token Index: 723

Text #3

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.4746. Min Act: -0.0000

Data Index: 21143 (C4 (Web Text))

Max Activating Token Index: 36

Text #4

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.4401. Min Act: -0.0000

Data Index: 86891 (C4 (Web Text))

Max Activating Token Index: 52

Text #5

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.5173. Min Act: -0.0000

Data Index: 1062544 (C4 (Web Text))

Max Activating Token Index: 344

Text #6

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.3770. Min Act: -0.0000

Data Index: 97262 (C4 (Web Text))

Max Activating Token Index: 309

Text #7

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.3102. Min Act: -0.0000

Data Index: 484991 (C4 (Web Text))

Max Activating Token Index: 139

Text #8

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.2649. Min Act: -0.0000

Data Index: 383146 (C4 (Web Text))

Max Activating Token Index: 242

Text #9

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.3448. Min Act: -0.0000

Data Index: 585875 (C4 (Web Text))

Max Activating Token Index: 194

Text #10

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.2238. Min Act: -0.0000

Data Index: 1053798 (C4 (Web Text))

Max Activating Token Index: 553

Text #11

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.1058. Min Act: -0.0000

Data Index: 650764 (C4 (Web Text))

Max Activating Token Index: 946

Text #12

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.1635. Min Act: -0.0000

Data Index: 1389907 (Python Code)

Max Activating Token Index: 125

Text #13

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.2767. Min Act: -0.0000

Data Index: 583333 (C4 (Web Text))

Max Activating Token Index: 261

Text #14

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.0465. Min Act: -0.0000

Data Index: 1138110 (C4 (Web Text))

Max Activating Token Index: 734

Text #15

Max Range: 3.4627. Min Range: -3.4627

Max Act: 1.9729. Min Act: -0.0000

Data Index: 154423 (C4 (Web Text))

Max Activating Token Index: 189

Text #16

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.1235. Min Act: -0.0000

Data Index: 856646 (C4 (Web Text))

Max Activating Token Index: 35

Text #17

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.0150. Min Act: -0.0000

Data Index: 73836 (C4 (Web Text))

Max Activating Token Index: 325

Text #18

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.1591. Min Act: -0.0000

Data Index: 134705 (C4 (Web Text))

Max Activating Token Index: 269

Text #19

Max Range: 3.4627. Min Range: -3.4627

Max Act: 2.0279. Min Act: -0.0000

Data Index: 353830 (C4 (Web Text))

Max Activating Token Index: 31