Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 1350 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0324. Min Act: -0.0002

Data Index: 897924 (C4 (Web Text))

Max Activating Token Index: 14

Text #1

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0350. Min Act: -0.0001

Data Index: 509937 (C4 (Web Text))

Max Activating Token Index: 387

Text #2

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0336. Min Act: -0.0002

Data Index: 1598300 (Python Code)

Max Activating Token Index: 245

Text #3

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0321. Min Act: -0.0002

Data Index: 454951 (C4 (Web Text))

Max Activating Token Index: 80

Text #4

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0269. Min Act: -0.0002

Data Index: 565517 (C4 (Web Text))

Max Activating Token Index: 222

Text #5

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0313. Min Act: -0.0002

Data Index: 1368217 (Python Code)

Max Activating Token Index: 328

Text #6

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0302. Min Act: -0.0002

Data Index: 974829 (C4 (Web Text))

Max Activating Token Index: 11

Text #7

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0272. Min Act: -0.0000

Data Index: 349677 (C4 (Web Text))

Max Activating Token Index: 68

Text #8

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0262. Min Act: -0.0002

Data Index: 465357 (C4 (Web Text))

Max Activating Token Index: 34

Text #9

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0291. Min Act: -0.0002

Data Index: 1015381 (C4 (Web Text))

Max Activating Token Index: 989

Text #10

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0279. Min Act: -0.0001

Data Index: 344240 (C4 (Web Text))

Max Activating Token Index: 54

Text #11

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0273. Min Act: -0.0000

Data Index: 30005 (C4 (Web Text))

Max Activating Token Index: 247

Text #12

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0246. Min Act: -0.0002

Data Index: 176517 (C4 (Web Text))

Max Activating Token Index: 507

Text #13

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0254. Min Act: -0.0002

Data Index: 316816 (C4 (Web Text))

Max Activating Token Index: 140

Text #14

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0264. Min Act: -0.0001

Data Index: 955456 (C4 (Web Text))

Max Activating Token Index: 187

Text #15

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0259. Min Act: -0.0003

Data Index: 1595580 (Python Code)

Max Activating Token Index: 834

Text #16

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0269. Min Act: 0.0000

Data Index: 1268287 (C4 (Web Text))

Max Activating Token Index: 32

Text #17

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0261. Min Act: -0.0001

Data Index: 1298257 (C4 (Web Text))

Max Activating Token Index: 960

Text #18

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0249. Min Act: -0.0001

Data Index: 1050883 (C4 (Web Text))

Max Activating Token Index: 840

Text #19

Max Range: 0.0350. Min Range: -0.0350

Max Act: 0.0261. Min Act: -0.0002

Data Index: 123623 (C4 (Web Text))

Max Activating Token Index: 893