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

Dataset: The Pile

Neuron 38 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.2760. Min Range: -5.2760

Max Act: 5.2760. Min Act: -0.0000

Data Index: 107362 (The Pile)

Max Activating Token Index: 299

Text #1

Max Range: 5.2760. Min Range: -5.2760

Max Act: 5.0801. Min Act: -0.0000

Data Index: 1519014 (The Pile)

Max Activating Token Index: 265

Text #2

Max Range: 5.2760. Min Range: -5.2760

Max Act: 4.8806. Min Act: -0.0000

Data Index: 33850 (The Pile)

Max Activating Token Index: 107

Text #3

Max Range: 5.2760. Min Range: -5.2760

Max Act: 4.7719. Min Act: -0.0000

Data Index: 1948193 (The Pile)

Max Activating Token Index: 23

Text #4

Max Range: 5.2760. Min Range: -5.2760

Max Act: 4.7316. Min Act: -0.0000

Data Index: 1726495 (The Pile)

Max Activating Token Index: 323

Text #5

Max Range: 5.2760. Min Range: -5.2760

Max Act: 4.4972. Min Act: -0.0000

Data Index: 956582 (The Pile)

Max Activating Token Index: 103

Text #6

Max Range: 5.2760. Min Range: -5.2760

Max Act: 4.4484. Min Act: -0.0000

Data Index: 1131797 (The Pile)

Max Activating Token Index: 681

Text #7

Max Range: 5.2760. Min Range: -5.2760

Max Act: 4.0443. Min Act: -0.0000

Data Index: 1660519 (The Pile)

Max Activating Token Index: 608

Text #8

Max Range: 5.2760. Min Range: -5.2760

Max Act: 4.2905. Min Act: -0.0000

Data Index: 1358957 (The Pile)

Max Activating Token Index: 130

Text #9

Max Range: 5.2760. Min Range: -5.2760

Max Act: 4.1212. Min Act: -0.0000

Data Index: 1655065 (The Pile)

Max Activating Token Index: 430

Text #10

Max Range: 5.2760. Min Range: -5.2760

Max Act: 4.0052. Min Act: -0.0000

Data Index: 1736339 (The Pile)

Max Activating Token Index: 76

Text #11

Max Range: 5.2760. Min Range: -5.2760

Max Act: 3.9902. Min Act: -0.0000

Data Index: 206465 (The Pile)

Max Activating Token Index: 56

Text #12

Max Range: 5.2760. Min Range: -5.2760

Max Act: 3.9877. Min Act: -0.0000

Data Index: 726451 (The Pile)

Max Activating Token Index: 121

Text #13

Max Range: 5.2760. Min Range: -5.2760

Max Act: 3.7591. Min Act: -0.0000

Data Index: 136026 (The Pile)

Max Activating Token Index: 87

Text #14

Max Range: 5.2760. Min Range: -5.2760

Max Act: 3.7380. Min Act: -0.0000

Data Index: 1404787 (The Pile)

Max Activating Token Index: 215

Text #15

Max Range: 5.2760. Min Range: -5.2760

Max Act: 3.7011. Min Act: -0.0000

Data Index: 818097 (The Pile)

Max Activating Token Index: 213

Text #16

Max Range: 5.2760. Min Range: -5.2760

Max Act: 3.8017. Min Act: -0.0000

Data Index: 1905529 (The Pile)

Max Activating Token Index: 603

Text #17

Max Range: 5.2760. Min Range: -5.2760

Max Act: 3.4020. Min Act: -0.0000

Data Index: 976465 (The Pile)

Max Activating Token Index: 915

Text #18

Max Range: 5.2760. Min Range: -5.2760

Max Act: 3.6784. Min Act: -0.0000

Data Index: 761648 (The Pile)

Max Activating Token Index: 831

Text #19

Max Range: 5.2760. Min Range: -5.2760

Max Act: 3.1373. Min Act: -0.0000

Data Index: 858609 (The Pile)

Max Activating Token Index: 355