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

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

Neuron 1627 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: 0.0735. Min Range: -0.0735

Max Act: 0.0735. Min Act: -0.0001

Data Index: 171820 (C4 (Web Text))

Max Activating Token Index: 266

Text #1

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0729. Min Act: -0.0002

Data Index: 1618690 (Python Code)

Max Activating Token Index: 808

Text #2

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0697. Min Act: -0.0002

Data Index: 1395453 (Python Code)

Max Activating Token Index: 579

Text #3

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0670. Min Act: -0.0002

Data Index: 1182753 (C4 (Web Text))

Max Activating Token Index: 632

Text #4

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0668. Min Act: -0.0002

Data Index: 1537288 (Python Code)

Max Activating Token Index: 763

Text #5

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0650. Min Act: -0.0001

Data Index: 621932 (C4 (Web Text))

Max Activating Token Index: 844

Text #6

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0633. Min Act: -0.0002

Data Index: 874231 (C4 (Web Text))

Max Activating Token Index: 285

Text #7

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0680. Min Act: -0.0001

Data Index: 317430 (C4 (Web Text))

Max Activating Token Index: 1009

Text #8

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0633. Min Act: -0.0001

Data Index: 142404 (C4 (Web Text))

Max Activating Token Index: 808

Text #9

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0647. Min Act: -0.0002

Data Index: 45567 (C4 (Web Text))

Max Activating Token Index: 285

Text #10

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0639. Min Act: -0.0002

Data Index: 55449 (C4 (Web Text))

Max Activating Token Index: 510

Text #11

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0612. Min Act: -0.0002

Data Index: 945685 (C4 (Web Text))

Max Activating Token Index: 882

Text #12

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0645. Min Act: -0.0001

Data Index: 710146 (C4 (Web Text))

Max Activating Token Index: 794

Text #13

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0627. Min Act: -0.0001

Data Index: 345164 (C4 (Web Text))

Max Activating Token Index: 550

Text #14

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0638. Min Act: -0.0002

Data Index: 1218101 (C4 (Web Text))

Max Activating Token Index: 718

Text #15

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0617. Min Act: -0.0002

Data Index: 1574506 (Python Code)

Max Activating Token Index: 212

Text #16

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0634. Min Act: -0.0002

Data Index: 730463 (C4 (Web Text))

Max Activating Token Index: 561

Text #17

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0624. Min Act: -0.0002

Data Index: 710229 (C4 (Web Text))

Max Activating Token Index: 481

Text #18

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0621. Min Act: -0.0001

Data Index: 1324029 (C4 (Web Text))

Max Activating Token Index: 983

Text #19

Max Range: 0.0735. Min Range: -0.0735

Max Act: 0.0631. Min Act: -0.0002

Data Index: 716311 (C4 (Web Text))

Max Activating Token Index: 371