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

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

Neuron 32 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: 0.4351. Min Range: -0.4351

Max Act: 0.4351. Min Act: -0.0000

Data Index: 192830 (C4 (Web Text))

Max Activating Token Index: 63

Text #1

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.3875. Min Act: -0.0000

Data Index: 375195 (C4 (Web Text))

Max Activating Token Index: 541

Text #2

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.3921. Min Act: -0.0000

Data Index: 894049 (C4 (Web Text))

Max Activating Token Index: 238

Text #3

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.3503. Min Act: -0.0000

Data Index: 175987 (C4 (Web Text))

Max Activating Token Index: 321

Text #4

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.3094. Min Act: -0.0000

Data Index: 1008447 (C4 (Web Text))

Max Activating Token Index: 554

Text #5

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2737. Min Act: -0.0000

Data Index: 553013 (C4 (Web Text))

Max Activating Token Index: 157

Text #6

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2601. Min Act: -0.0000

Data Index: 794775 (C4 (Web Text))

Max Activating Token Index: 234

Text #7

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2906. Min Act: -0.0000

Data Index: 416220 (C4 (Web Text))

Max Activating Token Index: 410

Text #8

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.3368. Min Act: -0.0000

Data Index: 817234 (C4 (Web Text))

Max Activating Token Index: 533

Text #9

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2297. Min Act: -0.0000

Data Index: 700565 (C4 (Web Text))

Max Activating Token Index: 878

Text #10

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2627. Min Act: -0.0000

Data Index: 784972 (C4 (Web Text))

Max Activating Token Index: 107

Text #11

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2582. Min Act: -0.0000

Data Index: 952814 (C4 (Web Text))

Max Activating Token Index: 198

Text #12

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2211. Min Act: -0.0000

Data Index: 1531654 (Python Code)

Max Activating Token Index: 734

Text #13

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2336. Min Act: -0.0000

Data Index: 1496102 (Python Code)

Max Activating Token Index: 310

Text #14

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2430. Min Act: -0.0000

Data Index: 1273040 (C4 (Web Text))

Max Activating Token Index: 272

Text #15

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2481. Min Act: -0.0000

Data Index: 544027 (C4 (Web Text))

Max Activating Token Index: 606

Text #16

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2746. Min Act: -0.0000

Data Index: 1415107 (Python Code)

Max Activating Token Index: 391

Text #17

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2434. Min Act: -0.0000

Data Index: 1524553 (Python Code)

Max Activating Token Index: 268

Text #18

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2239. Min Act: -0.0000

Data Index: 816612 (C4 (Web Text))

Max Activating Token Index: 484

Text #19

Max Range: 0.4351. Min Range: -0.4351

Max Act: 0.2208. Min Act: -0.0000

Data Index: 963286 (C4 (Web Text))

Max Activating Token Index: 720