Model: SoLU Model: 2 Layers, 2048 Neurons per Layer

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

Neuron 650 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-2l')

Text #0

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0381. Min Act: -0.0003

Data Index: 1294248 (C4 (Web Text))

Max Activating Token Index: 908

Text #1

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0320. Min Act: -0.0003

Data Index: 1086839 (C4 (Web Text))

Max Activating Token Index: 835

Text #2

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0313. Min Act: -0.0003

Data Index: 562555 (C4 (Web Text))

Max Activating Token Index: 1019

Text #3

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0285. Min Act: -0.0003

Data Index: 1466110 (Python Code)

Max Activating Token Index: 654

Text #4

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0285. Min Act: -0.0003

Data Index: 1047744 (C4 (Web Text))

Max Activating Token Index: 796

Text #5

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0285. Min Act: -0.0003

Data Index: 1391778 (Python Code)

Max Activating Token Index: 693

Text #6

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0270. Min Act: -0.0003

Data Index: 1449647 (Python Code)

Max Activating Token Index: 743

Text #7

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0280. Min Act: -0.0003

Data Index: 1152201 (C4 (Web Text))

Max Activating Token Index: 540

Text #8

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0275. Min Act: -0.0003

Data Index: 426796 (C4 (Web Text))

Max Activating Token Index: 629

Text #9

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0268. Min Act: -0.0003

Data Index: 1552421 (Python Code)

Max Activating Token Index: 446

Text #10

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0263. Min Act: -0.0003

Data Index: 1568575 (Python Code)

Max Activating Token Index: 850

Text #11

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0266. Min Act: -0.0003

Data Index: 1567266 (Python Code)

Max Activating Token Index: 888

Text #12

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0262. Min Act: -0.0003

Data Index: 1650375 (Python Code)

Max Activating Token Index: 582

Text #13

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0258. Min Act: -0.0003

Data Index: 1650013 (Python Code)

Max Activating Token Index: 824

Text #14

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0256. Min Act: -0.0003

Data Index: 1399190 (Python Code)

Max Activating Token Index: 881

Text #15

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0261. Min Act: -0.0003

Data Index: 306923 (C4 (Web Text))

Max Activating Token Index: 692

Text #16

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0261. Min Act: -0.0003

Data Index: 403710 (C4 (Web Text))

Max Activating Token Index: 758

Text #17

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0255. Min Act: -0.0003

Data Index: 158558 (C4 (Web Text))

Max Activating Token Index: 310

Text #18

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0264. Min Act: -0.0002

Data Index: 1050517 (C4 (Web Text))

Max Activating Token Index: 824

Text #19

Max Range: 0.0381. Min Range: -0.0381

Max Act: 0.0257. Min Act: -0.0003

Data Index: 1578613 (Python Code)

Max Activating Token Index: 874