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

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

Neuron 72 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.1075. Min Act: -0.0005

Data Index: 823040 (C4 (Web Text))

Max Activating Token Index: 55

Text #1

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0885. Min Act: -0.0005

Data Index: 1102983 (C4 (Web Text))

Max Activating Token Index: 938

Text #2

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0867. Min Act: -0.0005

Data Index: 506885 (C4 (Web Text))

Max Activating Token Index: 27

Text #3

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0817. Min Act: -0.0004

Data Index: 822524 (C4 (Web Text))

Max Activating Token Index: 701

Text #4

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0745. Min Act: -0.0004

Data Index: 867962 (C4 (Web Text))

Max Activating Token Index: 768

Text #5

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0769. Min Act: -0.0005

Data Index: 1322631 (C4 (Web Text))

Max Activating Token Index: 224

Text #6

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0770. Min Act: -0.0004

Data Index: 113732 (C4 (Web Text))

Max Activating Token Index: 557

Text #7

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0730. Min Act: -0.0005

Data Index: 674701 (C4 (Web Text))

Max Activating Token Index: 930

Text #8

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0753. Min Act: -0.0005

Data Index: 669513 (C4 (Web Text))

Max Activating Token Index: 101

Text #9

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0752. Min Act: -0.0004

Data Index: 399796 (C4 (Web Text))

Max Activating Token Index: 674

Text #10

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0772. Min Act: -0.0005

Data Index: 257894 (C4 (Web Text))

Max Activating Token Index: 984

Text #11

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0720. Min Act: -0.0004

Data Index: 149178 (C4 (Web Text))

Max Activating Token Index: 409

Text #12

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0687. Min Act: -0.0005

Data Index: 585934 (C4 (Web Text))

Max Activating Token Index: 780

Text #13

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0743. Min Act: -0.0005

Data Index: 788327 (C4 (Web Text))

Max Activating Token Index: 730

Text #14

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0736. Min Act: -0.0005

Data Index: 319935 (C4 (Web Text))

Max Activating Token Index: 204

Text #15

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0704. Min Act: -0.0005

Data Index: 738586 (C4 (Web Text))

Max Activating Token Index: 828

Text #16

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0692. Min Act: -0.0005

Data Index: 1061675 (C4 (Web Text))

Max Activating Token Index: 248

Text #17

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0736. Min Act: -0.0005

Data Index: 1348395 (C4 (Web Text))

Max Activating Token Index: 705

Text #18

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0719. Min Act: -0.0004

Data Index: 1057659 (C4 (Web Text))

Max Activating Token Index: 888

Text #19

Max Range: 0.1075. Min Range: -0.1075

Max Act: 0.0697. Min Act: -0.0004

Data Index: 782979 (C4 (Web Text))

Max Activating Token Index: 413