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

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

Neuron 1668 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: 0.0648. Min Range: -0.0648

Max Act: 0.0648. Min Act: -0.0003

Data Index: 1530558 (Python Code)

Max Activating Token Index: 900

Text #1

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0619. Min Act: -0.0002

Data Index: 555918 (C4 (Web Text))

Max Activating Token Index: 898

Text #2

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0630. Min Act: -0.0003

Data Index: 1484956 (Python Code)

Max Activating Token Index: 1019

Text #3

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0622. Min Act: -0.0003

Data Index: 1382282 (Python Code)

Max Activating Token Index: 929

Text #4

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0604. Min Act: -0.0003

Data Index: 146139 (C4 (Web Text))

Max Activating Token Index: 805

Text #5

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0616. Min Act: -0.0003

Data Index: 1531840 (Python Code)

Max Activating Token Index: 673

Text #6

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0608. Min Act: -0.0003

Data Index: 1393389 (Python Code)

Max Activating Token Index: 901

Text #7

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0604. Min Act: -0.0003

Data Index: 625048 (C4 (Web Text))

Max Activating Token Index: 843

Text #8

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0619. Min Act: -0.0003

Data Index: 1524746 (Python Code)

Max Activating Token Index: 803

Text #9

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0575. Min Act: -0.0003

Data Index: 26122 (C4 (Web Text))

Max Activating Token Index: 856

Text #10

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0583. Min Act: -0.0003

Data Index: 893861 (C4 (Web Text))

Max Activating Token Index: 354

Text #11

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0640. Min Act: -0.0003

Data Index: 1392497 (Python Code)

Max Activating Token Index: 1000

Text #12

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0572. Min Act: -0.0002

Data Index: 1068614 (C4 (Web Text))

Max Activating Token Index: 787

Text #13

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0603. Min Act: -0.0003

Data Index: 47948 (C4 (Web Text))

Max Activating Token Index: 684

Text #14

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0628. Min Act: -0.0003

Data Index: 1364937 (Python Code)

Max Activating Token Index: 988

Text #15

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0555. Min Act: -0.0003

Data Index: 106103 (C4 (Web Text))

Max Activating Token Index: 831

Text #16

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0552. Min Act: -0.0003

Data Index: 326497 (C4 (Web Text))

Max Activating Token Index: 420

Text #17

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0573. Min Act: -0.0003

Data Index: 425259 (C4 (Web Text))

Max Activating Token Index: 912

Text #18

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0548. Min Act: -0.0003

Data Index: 1440397 (Python Code)

Max Activating Token Index: 988

Text #19

Max Range: 0.0648. Min Range: -0.0648

Max Act: 0.0521. Min Act: -0.0003

Data Index: 1505648 (Python Code)

Max Activating Token Index: 496