Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 1374 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0766. Min Act: -0.0003

Data Index: 132878 (C4 (Web Text))

Max Activating Token Index: 602

Text #1

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0684. Min Act: -0.0003

Data Index: 821879 (C4 (Web Text))

Max Activating Token Index: 12

Text #2

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0697. Min Act: -0.0002

Data Index: 1491879 (Python Code)

Max Activating Token Index: 996

Text #3

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0677. Min Act: -0.0003

Data Index: 1073841 (C4 (Web Text))

Max Activating Token Index: 285

Text #4

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0637. Min Act: -0.0002

Data Index: 899877 (C4 (Web Text))

Max Activating Token Index: 915

Text #5

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0634. Min Act: -0.0003

Data Index: 670271 (C4 (Web Text))

Max Activating Token Index: 210

Text #6

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0627. Min Act: -0.0002

Data Index: 470231 (C4 (Web Text))

Max Activating Token Index: 639

Text #7

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0617. Min Act: -0.0003

Data Index: 1193420 (C4 (Web Text))

Max Activating Token Index: 76

Text #8

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0635. Min Act: -0.0003

Data Index: 768751 (C4 (Web Text))

Max Activating Token Index: 837

Text #9

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0607. Min Act: -0.0003

Data Index: 62238 (C4 (Web Text))

Max Activating Token Index: 82

Text #10

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0611. Min Act: -0.0003

Data Index: 922685 (C4 (Web Text))

Max Activating Token Index: 109

Text #11

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0589. Min Act: -0.0004

Data Index: 323400 (C4 (Web Text))

Max Activating Token Index: 687

Text #12

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0613. Min Act: -0.0004

Data Index: 464443 (C4 (Web Text))

Max Activating Token Index: 40

Text #13

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0585. Min Act: -0.0003

Data Index: 1549310 (Python Code)

Max Activating Token Index: 890

Text #14

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0584. Min Act: -0.0003

Data Index: 376131 (C4 (Web Text))

Max Activating Token Index: 357

Text #15

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0604. Min Act: -0.0002

Data Index: 1360581 (Python Code)

Max Activating Token Index: 852

Text #16

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0581. Min Act: -0.0003

Data Index: 1352628 (C4 (Web Text))

Max Activating Token Index: 888

Text #17

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0583. Min Act: -0.0003

Data Index: 296395 (C4 (Web Text))

Max Activating Token Index: 152

Text #18

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0568. Min Act: -0.0003

Data Index: 1552013 (Python Code)

Max Activating Token Index: 766

Text #19

Max Range: 0.0766. Min Range: -0.0766

Max Act: 0.0571. Min Act: -0.0003

Data Index: 332291 (C4 (Web Text))

Max Activating Token Index: 652