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

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

Neuron 3038 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.5737. Min Range: -2.5737

Max Act: 2.5737. Min Act: -0.0000

Data Index: 1275550 (C4 (Web Text))

Max Activating Token Index: 44

Text #1

Max Range: 2.5737. Min Range: -2.5737

Max Act: 2.3805. Min Act: 0.0000

Data Index: 1223204 (C4 (Web Text))

Max Activating Token Index: 459

Text #2

Max Range: 2.5737. Min Range: -2.5737

Max Act: 2.0276. Min Act: -0.0000

Data Index: 965611 (C4 (Web Text))

Max Activating Token Index: 396

Text #3

Max Range: 2.5737. Min Range: -2.5737

Max Act: 2.2391. Min Act: -0.0000

Data Index: 906316 (C4 (Web Text))

Max Activating Token Index: 240

Text #4

Max Range: 2.5737. Min Range: -2.5737

Max Act: 1.9648. Min Act: -0.0000

Data Index: 710615 (C4 (Web Text))

Max Activating Token Index: 884

Text #5

Max Range: 2.5737. Min Range: -2.5737

Max Act: 2.0204. Min Act: -0.0000

Data Index: 1144141 (C4 (Web Text))

Max Activating Token Index: 330

Text #6

Max Range: 2.5737. Min Range: -2.5737

Max Act: 1.7006. Min Act: -0.0000

Data Index: 504256 (C4 (Web Text))

Max Activating Token Index: 743

Text #7

Max Range: 2.5737. Min Range: -2.5737

Max Act: 1.5925. Min Act: -0.0000

Data Index: 1142583 (C4 (Web Text))

Max Activating Token Index: 61

Text #8

Max Range: 2.5737. Min Range: -2.5737

Max Act: 1.4455. Min Act: -0.0000

Data Index: 224975 (C4 (Web Text))

Max Activating Token Index: 352

Text #9

Max Range: 2.5737. Min Range: -2.5737

Max Act: 1.3938. Min Act: 0.0000

Data Index: 1639102 (Python Code)

Max Activating Token Index: 586

Text #10

Max Range: 2.5737. Min Range: -2.5737

Max Act: 1.2333. Min Act: -0.0000

Data Index: 96065 (C4 (Web Text))

Max Activating Token Index: 229

Text #11

Max Range: 2.5737. Min Range: -2.5737

Max Act: 1.2914. Min Act: -0.0000

Data Index: 1117520 (C4 (Web Text))

Max Activating Token Index: 65

Text #12

Max Range: 2.5737. Min Range: -2.5737

Max Act: 1.2446. Min Act: 0.0000

Data Index: 1400490 (Python Code)

Max Activating Token Index: 43

Text #13

Max Range: 2.5737. Min Range: -2.5737

Max Act: 1.1404. Min Act: 0.0000

Data Index: 1576746 (Python Code)

Max Activating Token Index: 979

Text #14

Max Range: 2.5737. Min Range: -2.5737

Max Act: 1.0751. Min Act: -0.0000

Data Index: 815199 (C4 (Web Text))

Max Activating Token Index: 505

Text #15

Max Range: 2.5737. Min Range: -2.5737

Max Act: 1.0164. Min Act: -0.0000

Data Index: 1454139 (Python Code)

Max Activating Token Index: 934

Text #16

Max Range: 2.5737. Min Range: -2.5737

Max Act: 1.0309. Min Act: 0.0000

Data Index: 1545397 (Python Code)

Max Activating Token Index: 942

Text #17

Max Range: 2.5737. Min Range: -2.5737

Max Act: 0.9943. Min Act: -0.0000

Data Index: 1136946 (C4 (Web Text))

Max Activating Token Index: 769

Text #18

Max Range: 2.5737. Min Range: -2.5737

Max Act: 0.9744. Min Act: -0.0000

Data Index: 1171994 (C4 (Web Text))

Max Activating Token Index: 717

Text #19

Max Range: 2.5737. Min Range: -2.5737

Max Act: 0.8871. Min Act: -0.0000

Data Index: 1532229 (Python Code)

Max Activating Token Index: 781