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

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

Neuron 47 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: 0.0917. Min Range: -0.0917

Max Act: 0.0917. Min Act: -0.0001

Data Index: 791546 (C4 (Web Text))

Max Activating Token Index: 727

Text #1

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0664. Min Act: -0.0001

Data Index: 809786 (C4 (Web Text))

Max Activating Token Index: 868

Text #2

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0643. Min Act: -0.0001

Data Index: 1016984 (C4 (Web Text))

Max Activating Token Index: 451

Text #3

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0578. Min Act: -0.0001

Data Index: 666831 (C4 (Web Text))

Max Activating Token Index: 716

Text #4

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0556. Min Act: -0.0002

Data Index: 1516321 (Python Code)

Max Activating Token Index: 935

Text #5

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0522. Min Act: -0.0001

Data Index: 1491760 (Python Code)

Max Activating Token Index: 958

Text #6

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0568. Min Act: -0.0002

Data Index: 1648678 (Python Code)

Max Activating Token Index: 465

Text #7

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0546. Min Act: -0.0000

Data Index: 700676 (C4 (Web Text))

Max Activating Token Index: 856

Text #8

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0571. Min Act: -0.0001

Data Index: 381199 (C4 (Web Text))

Max Activating Token Index: 856

Text #9

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0532. Min Act: -0.0000

Data Index: 1119480 (C4 (Web Text))

Max Activating Token Index: 511

Text #10

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0524. Min Act: -0.0001

Data Index: 1432909 (Python Code)

Max Activating Token Index: 987

Text #11

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0536. Min Act: -0.0000

Data Index: 1285020 (C4 (Web Text))

Max Activating Token Index: 1004

Text #12

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0505. Min Act: -0.0002

Data Index: 390792 (C4 (Web Text))

Max Activating Token Index: 498

Text #13

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0542. Min Act: -0.0002

Data Index: 1530683 (Python Code)

Max Activating Token Index: 998

Text #14

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0540. Min Act: -0.0000

Data Index: 1087925 (C4 (Web Text))

Max Activating Token Index: 715

Text #15

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0528. Min Act: -0.0002

Data Index: 64631 (C4 (Web Text))

Max Activating Token Index: 750

Text #16

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0518. Min Act: -0.0001

Data Index: 72065 (C4 (Web Text))

Max Activating Token Index: 629

Text #17

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0528. Min Act: -0.0001

Data Index: 1522794 (Python Code)

Max Activating Token Index: 154

Text #18

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0491. Min Act: -0.0000

Data Index: 72105 (C4 (Web Text))

Max Activating Token Index: 264

Text #19

Max Range: 0.0917. Min Range: -0.0917

Max Act: 0.0530. Min Act: -0.0001

Data Index: 1250222 (C4 (Web Text))

Max Activating Token Index: 956