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

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

Neuron 2848 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.7395. Min Act: -0.0000

Data Index: 119726 (C4 (Web Text))

Max Activating Token Index: 822

Text #1

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.6750. Min Act: -0.0000

Data Index: 609062 (C4 (Web Text))

Max Activating Token Index: 928

Text #2

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.5210. Min Act: -0.0000

Data Index: 169823 (C4 (Web Text))

Max Activating Token Index: 281

Text #3

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.4234. Min Act: -0.0000

Data Index: 1008027 (C4 (Web Text))

Max Activating Token Index: 406

Text #4

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.1919. Min Act: -0.0000

Data Index: 851253 (C4 (Web Text))

Max Activating Token Index: 579

Text #5

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.3381. Min Act: -0.0000

Data Index: 795135 (C4 (Web Text))

Max Activating Token Index: 972

Text #6

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.2243. Min Act: -0.0000

Data Index: 667163 (C4 (Web Text))

Max Activating Token Index: 276

Text #7

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.1753. Min Act: -0.0000

Data Index: 600269 (C4 (Web Text))

Max Activating Token Index: 201

Text #8

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.0971. Min Act: -0.0000

Data Index: 833157 (C4 (Web Text))

Max Activating Token Index: 291

Text #9

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.0961. Min Act: -0.0000

Data Index: 1395961 (Python Code)

Max Activating Token Index: 129

Text #10

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.1797. Min Act: -0.0000

Data Index: 601903 (C4 (Web Text))

Max Activating Token Index: 501

Text #11

Max Range: 2.7395. Min Range: -2.7395

Max Act: 1.9331. Min Act: -0.0000

Data Index: 1226409 (C4 (Web Text))

Max Activating Token Index: 138

Text #12

Max Range: 2.7395. Min Range: -2.7395

Max Act: 1.9310. Min Act: -0.0000

Data Index: 505988 (C4 (Web Text))

Max Activating Token Index: 108

Text #13

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.1209. Min Act: -0.0000

Data Index: 1054692 (C4 (Web Text))

Max Activating Token Index: 478

Text #14

Max Range: 2.7395. Min Range: -2.7395

Max Act: 1.9206. Min Act: -0.0000

Data Index: 1332554 (C4 (Web Text))

Max Activating Token Index: 940

Text #15

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.0074. Min Act: -0.0000

Data Index: 323549 (C4 (Web Text))

Max Activating Token Index: 908

Text #16

Max Range: 2.7395. Min Range: -2.7395

Max Act: 1.9136. Min Act: -0.0000

Data Index: 335375 (C4 (Web Text))

Max Activating Token Index: 944

Text #17

Max Range: 2.7395. Min Range: -2.7395

Max Act: 1.9400. Min Act: -0.0001

Data Index: 1399870 (Python Code)

Max Activating Token Index: 249

Text #18

Max Range: 2.7395. Min Range: -2.7395

Max Act: 2.0029. Min Act: -0.0000

Data Index: 1204453 (C4 (Web Text))

Max Activating Token Index: 710

Text #19

Max Range: 2.7395. Min Range: -2.7395

Max Act: 1.9702. Min Act: -0.0000

Data Index: 576591 (C4 (Web Text))

Max Activating Token Index: 274