Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 5610 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 3.8826. Min Range: -3.8826

Max Act: 3.8826. Min Act: -0.0000

Data Index: 706834 (C4 (Web Text))

Max Activating Token Index: 886

Text #1

Max Range: 3.8826. Min Range: -3.8826

Max Act: 3.8135. Min Act: -0.0000

Data Index: 195895 (C4 (Web Text))

Max Activating Token Index: 649

Text #2

Max Range: 3.8826. Min Range: -3.8826

Max Act: 3.5561. Min Act: -0.0000

Data Index: 1422529 (Python Code)

Max Activating Token Index: 783

Text #3

Max Range: 3.8826. Min Range: -3.8826

Max Act: 3.4010. Min Act: -0.0000

Data Index: 1532264 (Python Code)

Max Activating Token Index: 837

Text #4

Max Range: 3.8826. Min Range: -3.8826

Max Act: 3.4362. Min Act: -0.0000

Data Index: 1644683 (Python Code)

Max Activating Token Index: 865

Text #5

Max Range: 3.8826. Min Range: -3.8826

Max Act: 3.0890. Min Act: -0.0000

Data Index: 1552629 (Python Code)

Max Activating Token Index: 659

Text #6

Max Range: 3.8826. Min Range: -3.8826

Max Act: 3.0298. Min Act: -0.0000

Data Index: 1631900 (Python Code)

Max Activating Token Index: 378

Text #7

Max Range: 3.8826. Min Range: -3.8826

Max Act: 2.9418. Min Act: -0.0000

Data Index: 728586 (C4 (Web Text))

Max Activating Token Index: 24

Text #8

Max Range: 3.8826. Min Range: -3.8826

Max Act: 2.9083. Min Act: -0.0000

Data Index: 1023060 (C4 (Web Text))

Max Activating Token Index: 570

Text #9

Max Range: 3.8826. Min Range: -3.8826

Max Act: 2.8219. Min Act: -0.0000

Data Index: 1504082 (Python Code)

Max Activating Token Index: 331

Text #10

Max Range: 3.8826. Min Range: -3.8826

Max Act: 3.0550. Min Act: -0.0000

Data Index: 1441454 (Python Code)

Max Activating Token Index: 845

Text #11

Max Range: 3.8826. Min Range: -3.8826

Max Act: 2.8173. Min Act: -0.0000

Data Index: 1625945 (Python Code)

Max Activating Token Index: 410

Text #12

Max Range: 3.8826. Min Range: -3.8826

Max Act: 2.8054. Min Act: -0.0000

Data Index: 1400192 (Python Code)

Max Activating Token Index: 609

Text #13

Max Range: 3.8826. Min Range: -3.8826

Max Act: 2.7011. Min Act: -0.0000

Data Index: 1401130 (Python Code)

Max Activating Token Index: 552

Text #14

Max Range: 3.8826. Min Range: -3.8826

Max Act: 2.5501. Min Act: -0.0000

Data Index: 1615188 (Python Code)

Max Activating Token Index: 925

Text #15

Max Range: 3.8826. Min Range: -3.8826

Max Act: 2.6455. Min Act: -0.0000

Data Index: 1641162 (Python Code)

Max Activating Token Index: 521

Text #16

Max Range: 3.8826. Min Range: -3.8826

Max Act: 2.7286. Min Act: -0.0000

Data Index: 1453616 (Python Code)

Max Activating Token Index: 350

Text #17

Max Range: 3.8826. Min Range: -3.8826

Max Act: 2.5665. Min Act: -0.0000

Data Index: 1543563 (Python Code)

Max Activating Token Index: 929

Text #18

Max Range: 3.8826. Min Range: -3.8826

Max Act: 2.3538. Min Act: -0.0000

Data Index: 1557627 (Python Code)

Max Activating Token Index: 239

Text #19

Max Range: 3.8826. Min Range: -3.8826

Max Act: 2.7055. Min Act: -0.0000

Data Index: 348416 (C4 (Web Text))

Max Activating Token Index: 571