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

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

Neuron 4321 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.9541. Min Act: -0.0000

Data Index: 1396176 (Python Code)

Max Activating Token Index: 715

Text #1

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.8573. Min Act: -0.0000

Data Index: 1521197 (Python Code)

Max Activating Token Index: 972

Text #2

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.7299. Min Act: -0.0000

Data Index: 1017957 (C4 (Web Text))

Max Activating Token Index: 397

Text #3

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.5412. Min Act: -0.0000

Data Index: 1395924 (Python Code)

Max Activating Token Index: 859

Text #4

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.2226. Min Act: -0.0000

Data Index: 1354618 (C4 (Web Text))

Max Activating Token Index: 963

Text #5

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.2353. Min Act: -0.0000

Data Index: 1540152 (Python Code)

Max Activating Token Index: 973

Text #6

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.3317. Min Act: -0.0000

Data Index: 942903 (C4 (Web Text))

Max Activating Token Index: 269

Text #7

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.1444. Min Act: -0.0000

Data Index: 755041 (C4 (Web Text))

Max Activating Token Index: 27

Text #8

Max Range: 2.9541. Min Range: -2.9541

Max Act: 1.9690. Min Act: -0.0000

Data Index: 1532314 (Python Code)

Max Activating Token Index: 232

Text #9

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.0377. Min Act: -0.0000

Data Index: 1562738 (Python Code)

Max Activating Token Index: 814

Text #10

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.0034. Min Act: -0.0000

Data Index: 1547403 (Python Code)

Max Activating Token Index: 809

Text #11

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.0647. Min Act: -0.0000

Data Index: 1408963 (Python Code)

Max Activating Token Index: 877

Text #12

Max Range: 2.9541. Min Range: -2.9541

Max Act: 1.8405. Min Act: -0.0000

Data Index: 1576164 (Python Code)

Max Activating Token Index: 64

Text #13

Max Range: 2.9541. Min Range: -2.9541

Max Act: 1.9224. Min Act: -0.0000

Data Index: 649005 (C4 (Web Text))

Max Activating Token Index: 15

Text #14

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.0229. Min Act: -0.0000

Data Index: 1184190 (C4 (Web Text))

Max Activating Token Index: 472

Text #15

Max Range: 2.9541. Min Range: -2.9541

Max Act: 1.9147. Min Act: -0.0000

Data Index: 1480317 (Python Code)

Max Activating Token Index: 224

Text #16

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.0094. Min Act: -0.0000

Data Index: 1061447 (C4 (Web Text))

Max Activating Token Index: 175

Text #17

Max Range: 2.9541. Min Range: -2.9541

Max Act: 1.9145. Min Act: -0.0000

Data Index: 1447955 (Python Code)

Max Activating Token Index: 90

Text #18

Max Range: 2.9541. Min Range: -2.9541

Max Act: 2.0055. Min Act: -0.0000

Data Index: 1499107 (Python Code)

Max Activating Token Index: 395

Text #19

Max Range: 2.9541. Min Range: -2.9541

Max Act: 1.8654. Min Act: -0.0000

Data Index: 841368 (C4 (Web Text))

Max Activating Token Index: 148