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

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

Neuron 5089 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.9940. Min Act: -0.0002

Data Index: 948001 (C4 (Web Text))

Max Activating Token Index: 697

Text #1

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.7357. Min Act: -0.0002

Data Index: 1532136 (Python Code)

Max Activating Token Index: 811

Text #2

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.6065. Min Act: -0.0002

Data Index: 1411431 (Python Code)

Max Activating Token Index: 754

Text #3

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.4807. Min Act: -0.0002

Data Index: 1396972 (Python Code)

Max Activating Token Index: 794

Text #4

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.2359. Min Act: -0.0002

Data Index: 624379 (C4 (Web Text))

Max Activating Token Index: 541

Text #5

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.2317. Min Act: -0.0002

Data Index: 1554870 (Python Code)

Max Activating Token Index: 849

Text #6

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.3548. Min Act: -0.0002

Data Index: 1633236 (Python Code)

Max Activating Token Index: 636

Text #7

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.3540. Min Act: -0.0002

Data Index: 1620447 (Python Code)

Max Activating Token Index: 846

Text #8

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.1098. Min Act: -0.0001

Data Index: 958764 (C4 (Web Text))

Max Activating Token Index: 560

Text #9

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.2353. Min Act: -0.0002

Data Index: 1285298 (C4 (Web Text))

Max Activating Token Index: 726

Text #10

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.2333. Min Act: -0.0002

Data Index: 631788 (C4 (Web Text))

Max Activating Token Index: 836

Text #11

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.2211. Min Act: -0.0002

Data Index: 1624659 (Python Code)

Max Activating Token Index: 720

Text #12

Max Range: 27.9940. Min Range: -27.9940

Max Act: 26.9888. Min Act: -0.0002

Data Index: 1211074 (C4 (Web Text))

Max Activating Token Index: 787

Text #13

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.1075. Min Act: -0.0002

Data Index: 73479 (C4 (Web Text))

Max Activating Token Index: 561

Text #14

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.1097. Min Act: -0.0002

Data Index: 559714 (C4 (Web Text))

Max Activating Token Index: 680

Text #15

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.1084. Min Act: -0.0002

Data Index: 1599706 (Python Code)

Max Activating Token Index: 771

Text #16

Max Range: 27.9940. Min Range: -27.9940

Max Act: 27.1084. Min Act: -0.0002

Data Index: 906144 (C4 (Web Text))

Max Activating Token Index: 252

Text #17

Max Range: 27.9940. Min Range: -27.9940

Max Act: 26.8491. Min Act: -0.0002

Data Index: 1579928 (Python Code)

Max Activating Token Index: 763

Text #18

Max Range: 27.9940. Min Range: -27.9940

Max Act: 26.9896. Min Act: -0.0002

Data Index: 1177759 (C4 (Web Text))

Max Activating Token Index: 573

Text #19

Max Range: 27.9940. Min Range: -27.9940

Max Act: 26.9880. Min Act: -0.0002

Data Index: 1179783 (C4 (Web Text))

Max Activating Token Index: 317