Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 620 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 4.0794. Min Range: -4.0794

Max Act: 4.0512. Min Act: -0.0001

Data Index: 796128 (C4 (Web Text))

Max Activating Token Index: 508

Text #1

Max Range: 4.0794. Min Range: -4.0794

Max Act: 4.0794. Min Act: -0.0002

Data Index: 949704 (C4 (Web Text))

Max Activating Token Index: 546

Text #2

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.8300. Min Act: -0.0002

Data Index: 397985 (C4 (Web Text))

Max Activating Token Index: 453

Text #3

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.3448. Min Act: -0.0002

Data Index: 639390 (C4 (Web Text))

Max Activating Token Index: 494

Text #4

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.4282. Min Act: -0.0001

Data Index: 536520 (C4 (Web Text))

Max Activating Token Index: 558

Text #5

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.2714. Min Act: -0.0001

Data Index: 630513 (C4 (Web Text))

Max Activating Token Index: 678

Text #6

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.3679. Min Act: -0.0002

Data Index: 863623 (C4 (Web Text))

Max Activating Token Index: 321

Text #7

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.4469. Min Act: -0.0001

Data Index: 19253 (C4 (Web Text))

Max Activating Token Index: 519

Text #8

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.3199. Min Act: -0.0001

Data Index: 1217036 (C4 (Web Text))

Max Activating Token Index: 801

Text #9

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.1530. Min Act: -0.0001

Data Index: 727021 (C4 (Web Text))

Max Activating Token Index: 679

Text #10

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.2859. Min Act: -0.0001

Data Index: 1204267 (C4 (Web Text))

Max Activating Token Index: 551

Text #11

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.2051. Min Act: -0.0002

Data Index: 6139 (C4 (Web Text))

Max Activating Token Index: 478

Text #12

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.1841. Min Act: -0.0001

Data Index: 368551 (C4 (Web Text))

Max Activating Token Index: 600

Text #13

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.5097. Min Act: -0.0001

Data Index: 240390 (C4 (Web Text))

Max Activating Token Index: 604

Text #14

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.0602. Min Act: -0.0001

Data Index: 1277594 (C4 (Web Text))

Max Activating Token Index: 569

Text #15

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.1780. Min Act: -0.0002

Data Index: 828743 (C4 (Web Text))

Max Activating Token Index: 755

Text #16

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.1643. Min Act: -0.0001

Data Index: 179669 (C4 (Web Text))

Max Activating Token Index: 411

Text #17

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.1578. Min Act: -0.0001

Data Index: 135910 (C4 (Web Text))

Max Activating Token Index: 830

Text #18

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.3008. Min Act: -0.0002

Data Index: 982448 (C4 (Web Text))

Max Activating Token Index: 601

Text #19

Max Range: 4.0794. Min Range: -4.0794

Max Act: 3.1427. Min Act: -0.0001

Data Index: 293490 (C4 (Web Text))

Max Activating Token Index: 519