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

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

Neuron 3987 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0281. Min Act: -0.0000

Data Index: 567243 (C4 (Web Text))

Max Activating Token Index: 603

Text #1

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0314. Min Act: -0.0000

Data Index: 940903 (C4 (Web Text))

Max Activating Token Index: 711

Text #2

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0199. Min Act: -0.0001

Data Index: 842030 (C4 (Web Text))

Max Activating Token Index: 628

Text #3

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0288. Min Act: -0.0000

Data Index: 295965 (C4 (Web Text))

Max Activating Token Index: 497

Text #4

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0286. Min Act: -0.0000

Data Index: 480423 (C4 (Web Text))

Max Activating Token Index: 993

Text #5

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0279. Min Act: -0.0000

Data Index: 162433 (C4 (Web Text))

Max Activating Token Index: 330

Text #6

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0268. Min Act: -0.0000

Data Index: 469166 (C4 (Web Text))

Max Activating Token Index: 246

Text #7

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0246. Min Act: -0.0000

Data Index: 121227 (C4 (Web Text))

Max Activating Token Index: 1021

Text #8

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0258. Min Act: -0.0001

Data Index: 650664 (C4 (Web Text))

Max Activating Token Index: 768

Text #9

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0268. Min Act: -0.0000

Data Index: 1629302 (Python Code)

Max Activating Token Index: 66

Text #10

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0264. Min Act: -0.0000

Data Index: 1307748 (C4 (Web Text))

Max Activating Token Index: 973

Text #11

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0250. Min Act: -0.0000

Data Index: 1538644 (Python Code)

Max Activating Token Index: 323

Text #12

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0248. Min Act: -0.0001

Data Index: 980804 (C4 (Web Text))

Max Activating Token Index: 189

Text #13

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0245. Min Act: -0.0000

Data Index: 596687 (C4 (Web Text))

Max Activating Token Index: 505

Text #14

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0227. Min Act: -0.0000

Data Index: 627113 (C4 (Web Text))

Max Activating Token Index: 265

Text #15

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0233. Min Act: -0.0001

Data Index: 126914 (C4 (Web Text))

Max Activating Token Index: 942

Text #16

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0236. Min Act: -0.0000

Data Index: 1372844 (Python Code)

Max Activating Token Index: 847

Text #17

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0208. Min Act: -0.0000

Data Index: 1164472 (C4 (Web Text))

Max Activating Token Index: 260

Text #18

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0234. Min Act: -0.0000

Data Index: 457055 (C4 (Web Text))

Max Activating Token Index: 330

Text #19

Max Range: 0.0314. Min Range: -0.0314

Max Act: 0.0221. Min Act: -0.0000

Data Index: 280378 (C4 (Web Text))

Max Activating Token Index: 763