Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 764 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0637. Min Act: -0.0002

Data Index: 1230213 (C4 (Web Text))

Max Activating Token Index: 454

Text #1

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0559. Min Act: -0.0002

Data Index: 913803 (C4 (Web Text))

Max Activating Token Index: 763

Text #2

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0545. Min Act: -0.0002

Data Index: 1339157 (C4 (Web Text))

Max Activating Token Index: 618

Text #3

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0543. Min Act: -0.0001

Data Index: 601480 (C4 (Web Text))

Max Activating Token Index: 712

Text #4

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0551. Min Act: -0.0002

Data Index: 707093 (C4 (Web Text))

Max Activating Token Index: 775

Text #5

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0523. Min Act: -0.0002

Data Index: 1158992 (C4 (Web Text))

Max Activating Token Index: 181

Text #6

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0526. Min Act: -0.0002

Data Index: 1417179 (Python Code)

Max Activating Token Index: 856

Text #7

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0532. Min Act: -0.0002

Data Index: 1028189 (C4 (Web Text))

Max Activating Token Index: 219

Text #8

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0520. Min Act: -0.0002

Data Index: 1391542 (Python Code)

Max Activating Token Index: 748

Text #9

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0500. Min Act: -0.0002

Data Index: 1527334 (Python Code)

Max Activating Token Index: 155

Text #10

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0510. Min Act: -0.0002

Data Index: 1556376 (Python Code)

Max Activating Token Index: 171

Text #11

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0510. Min Act: -0.0002

Data Index: 1492098 (Python Code)

Max Activating Token Index: 639

Text #12

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0508. Min Act: -0.0002

Data Index: 1542914 (Python Code)

Max Activating Token Index: 56

Text #13

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0493. Min Act: -0.0002

Data Index: 555850 (C4 (Web Text))

Max Activating Token Index: 354

Text #14

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0493. Min Act: -0.0002

Data Index: 1043980 (C4 (Web Text))

Max Activating Token Index: 664

Text #15

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0504. Min Act: -0.0002

Data Index: 1461954 (Python Code)

Max Activating Token Index: 42

Text #16

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0494. Min Act: -0.0002

Data Index: 1403309 (Python Code)

Max Activating Token Index: 589

Text #17

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0501. Min Act: -0.0002

Data Index: 1426418 (Python Code)

Max Activating Token Index: 352

Text #18

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0512. Min Act: -0.0002

Data Index: 1606761 (Python Code)

Max Activating Token Index: 494

Text #19

Max Range: 0.0637. Min Range: -0.0637

Max Act: 0.0516. Min Act: -0.0002

Data Index: 1596768 (Python Code)

Max Activating Token Index: 943