Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 4151 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 5.8976. Min Range: -5.8976

Max Act: 5.8976. Min Act: -0.0000

Data Index: 1004075 (C4 (Web Text))

Max Activating Token Index: 65

Text #1

Max Range: 5.8976. Min Range: -5.8976

Max Act: 5.1866. Min Act: -0.0000

Data Index: 537437 (C4 (Web Text))

Max Activating Token Index: 103

Text #2

Max Range: 5.8976. Min Range: -5.8976

Max Act: 4.6203. Min Act: -0.0000

Data Index: 468563 (C4 (Web Text))

Max Activating Token Index: 103

Text #3

Max Range: 5.8976. Min Range: -5.8976

Max Act: 3.7491. Min Act: -0.0000

Data Index: 1582275 (Python Code)

Max Activating Token Index: 904

Text #4

Max Range: 5.8976. Min Range: -5.8976

Max Act: 3.6394. Min Act: -0.0000

Data Index: 1523657 (Python Code)

Max Activating Token Index: 574

Text #5

Max Range: 5.8976. Min Range: -5.8976

Max Act: 3.7888. Min Act: -0.0000

Data Index: 1635743 (Python Code)

Max Activating Token Index: 720

Text #6

Max Range: 5.8976. Min Range: -5.8976

Max Act: 3.4890. Min Act: -0.0000

Data Index: 953905 (C4 (Web Text))

Max Activating Token Index: 96

Text #7

Max Range: 5.8976. Min Range: -5.8976

Max Act: 3.1811. Min Act: -0.0000

Data Index: 1459338 (Python Code)

Max Activating Token Index: 617

Text #8

Max Range: 5.8976. Min Range: -5.8976

Max Act: 3.1853. Min Act: -0.0000

Data Index: 1565366 (Python Code)

Max Activating Token Index: 369

Text #9

Max Range: 5.8976. Min Range: -5.8976

Max Act: 3.1879. Min Act: -0.0000

Data Index: 1488538 (Python Code)

Max Activating Token Index: 480

Text #10

Max Range: 5.8976. Min Range: -5.8976

Max Act: 3.0657. Min Act: -0.0000

Data Index: 1313687 (C4 (Web Text))

Max Activating Token Index: 355

Text #11

Max Range: 5.8976. Min Range: -5.8976

Max Act: 2.9794. Min Act: -0.0000

Data Index: 1501961 (Python Code)

Max Activating Token Index: 536

Text #12

Max Range: 5.8976. Min Range: -5.8976

Max Act: 3.0659. Min Act: -0.0000

Data Index: 7582 (C4 (Web Text))

Max Activating Token Index: 911

Text #13

Max Range: 5.8976. Min Range: -5.8976

Max Act: 2.5463. Min Act: -0.0000

Data Index: 581222 (C4 (Web Text))

Max Activating Token Index: 994

Text #14

Max Range: 5.8976. Min Range: -5.8976

Max Act: 2.7038. Min Act: -0.0000

Data Index: 1590151 (Python Code)

Max Activating Token Index: 184

Text #15

Max Range: 5.8976. Min Range: -5.8976

Max Act: 2.5898. Min Act: -0.0000

Data Index: 1333978 (C4 (Web Text))

Max Activating Token Index: 516

Text #16

Max Range: 5.8976. Min Range: -5.8976

Max Act: 2.4240. Min Act: -0.0000

Data Index: 876701 (C4 (Web Text))

Max Activating Token Index: 195

Text #17

Max Range: 5.8976. Min Range: -5.8976

Max Act: 2.3385. Min Act: -0.0000

Data Index: 1340624 (C4 (Web Text))

Max Activating Token Index: 358

Text #18

Max Range: 5.8976. Min Range: -5.8976

Max Act: 2.4856. Min Act: -0.0001

Data Index: 1402034 (Python Code)

Max Activating Token Index: 780

Text #19

Max Range: 5.8976. Min Range: -5.8976

Max Act: 2.4454. Min Act: -0.0000

Data Index: 299438 (C4 (Web Text))

Max Activating Token Index: 85