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

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

Neuron 4712 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 4.8833. Min Range: -4.8833

Max Act: 4.7580. Min Act: -0.0000

Data Index: 72371 (C4 (Web Text))

Max Activating Token Index: 482

Text #1

Max Range: 4.8833. Min Range: -4.8833

Max Act: 4.8833. Min Act: -0.0000

Data Index: 956335 (C4 (Web Text))

Max Activating Token Index: 616

Text #2

Max Range: 4.8833. Min Range: -4.8833

Max Act: 4.1963. Min Act: -0.0000

Data Index: 895770 (C4 (Web Text))

Max Activating Token Index: 314

Text #3

Max Range: 4.8833. Min Range: -4.8833

Max Act: 4.2794. Min Act: -0.0000

Data Index: 1211575 (C4 (Web Text))

Max Activating Token Index: 894

Text #4

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.9736. Min Act: -0.0000

Data Index: 489616 (C4 (Web Text))

Max Activating Token Index: 222

Text #5

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.9195. Min Act: -0.0000

Data Index: 257497 (C4 (Web Text))

Max Activating Token Index: 82

Text #6

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.5501. Min Act: -0.0000

Data Index: 784513 (C4 (Web Text))

Max Activating Token Index: 198

Text #7

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.5510. Min Act: -0.0000

Data Index: 500714 (C4 (Web Text))

Max Activating Token Index: 770

Text #8

Max Range: 4.8833. Min Range: -4.8833

Max Act: 4.0045. Min Act: -0.0000

Data Index: 1338287 (C4 (Web Text))

Max Activating Token Index: 711

Text #9

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.4465. Min Act: -0.0000

Data Index: 885074 (C4 (Web Text))

Max Activating Token Index: 611

Text #10

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.0339. Min Act: -0.0000

Data Index: 312537 (C4 (Web Text))

Max Activating Token Index: 879

Text #11

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.1504. Min Act: -0.0000

Data Index: 1558368 (Python Code)

Max Activating Token Index: 785

Text #12

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.2792. Min Act: -0.0000

Data Index: 646383 (C4 (Web Text))

Max Activating Token Index: 259

Text #13

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.4266. Min Act: -0.0000

Data Index: 998441 (C4 (Web Text))

Max Activating Token Index: 855

Text #14

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.0059. Min Act: -0.0000

Data Index: 624067 (C4 (Web Text))

Max Activating Token Index: 933

Text #15

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.0383. Min Act: -0.0000

Data Index: 1566861 (Python Code)

Max Activating Token Index: 325

Text #16

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.0318. Min Act: -0.0000

Data Index: 309537 (C4 (Web Text))

Max Activating Token Index: 255

Text #17

Max Range: 4.8833. Min Range: -4.8833

Max Act: 3.0272. Min Act: -0.0000

Data Index: 186766 (C4 (Web Text))

Max Activating Token Index: 695

Text #18

Max Range: 4.8833. Min Range: -4.8833

Max Act: 2.9297. Min Act: -0.0000

Data Index: 682355 (C4 (Web Text))

Max Activating Token Index: 512

Text #19

Max Range: 4.8833. Min Range: -4.8833

Max Act: 2.8964. Min Act: -0.0000

Data Index: 1046744 (C4 (Web Text))

Max Activating Token Index: 983