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

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

Neuron 4768 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0440. Min Act: -0.0000

Data Index: 704302 (C4 (Web Text))

Max Activating Token Index: 43

Text #1

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0392. Min Act: -0.0001

Data Index: 1545207 (Python Code)

Max Activating Token Index: 76

Text #2

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0413. Min Act: -0.0001

Data Index: 1442538 (Python Code)

Max Activating Token Index: 388

Text #3

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0403. Min Act: -0.0001

Data Index: 1651148 (Python Code)

Max Activating Token Index: 903

Text #4

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0402. Min Act: -0.0000

Data Index: 1624578 (Python Code)

Max Activating Token Index: 503

Text #5

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0374. Min Act: -0.0000

Data Index: 1169743 (C4 (Web Text))

Max Activating Token Index: 656

Text #6

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0375. Min Act: -0.0000

Data Index: 878225 (C4 (Web Text))

Max Activating Token Index: 345

Text #7

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0371. Min Act: -0.0000

Data Index: 702365 (C4 (Web Text))

Max Activating Token Index: 794

Text #8

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0355. Min Act: -0.0000

Data Index: 1022414 (C4 (Web Text))

Max Activating Token Index: 450

Text #9

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0379. Min Act: -0.0000

Data Index: 1633622 (Python Code)

Max Activating Token Index: 790

Text #10

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0361. Min Act: -0.0000

Data Index: 683672 (C4 (Web Text))

Max Activating Token Index: 367

Text #11

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0370. Min Act: -0.0000

Data Index: 190906 (C4 (Web Text))

Max Activating Token Index: 876

Text #12

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0338. Min Act: -0.0000

Data Index: 463553 (C4 (Web Text))

Max Activating Token Index: 341

Text #13

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0379. Min Act: -0.0000

Data Index: 471874 (C4 (Web Text))

Max Activating Token Index: 647

Text #14

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0377. Min Act: -0.0001

Data Index: 155751 (C4 (Web Text))

Max Activating Token Index: 765

Text #15

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0360. Min Act: -0.0000

Data Index: 838474 (C4 (Web Text))

Max Activating Token Index: 745

Text #16

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0372. Min Act: -0.0000

Data Index: 408092 (C4 (Web Text))

Max Activating Token Index: 514

Text #17

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0333. Min Act: -0.0000

Data Index: 843413 (C4 (Web Text))

Max Activating Token Index: 513

Text #18

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0327. Min Act: -0.0001

Data Index: 1492366 (Python Code)

Max Activating Token Index: 511

Text #19

Max Range: 0.0440. Min Range: -0.0440

Max Act: 0.0318. Min Act: -0.0000

Data Index: 541436 (C4 (Web Text))

Max Activating Token Index: 266