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

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

Neuron 4830 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.7236. Min Act: -0.0001

Data Index: 1318040 (C4 (Web Text))

Max Activating Token Index: 760

Text #1

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.7404. Min Act: -0.0001

Data Index: 1410371 (Python Code)

Max Activating Token Index: 835

Text #2

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.7128. Min Act: -0.0001

Data Index: 109399 (C4 (Web Text))

Max Activating Token Index: 961

Text #3

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6917. Min Act: -0.0001

Data Index: 770647 (C4 (Web Text))

Max Activating Token Index: 839

Text #4

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6915. Min Act: -0.0001

Data Index: 1153852 (C4 (Web Text))

Max Activating Token Index: 356

Text #5

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.7427. Min Act: -0.0001

Data Index: 56493 (C4 (Web Text))

Max Activating Token Index: 702

Text #6

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6729. Min Act: -0.0001

Data Index: 997932 (C4 (Web Text))

Max Activating Token Index: 947

Text #7

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6489. Min Act: -0.0001

Data Index: 1173988 (C4 (Web Text))

Max Activating Token Index: 12

Text #8

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6385. Min Act: -0.0001

Data Index: 994720 (C4 (Web Text))

Max Activating Token Index: 498

Text #9

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6796. Min Act: -0.0001

Data Index: 1249465 (C4 (Web Text))

Max Activating Token Index: 743

Text #10

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6493. Min Act: -0.0001

Data Index: 1140882 (C4 (Web Text))

Max Activating Token Index: 545

Text #11

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6316. Min Act: -0.0001

Data Index: 1032172 (C4 (Web Text))

Max Activating Token Index: 415

Text #12

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6475. Min Act: -0.0001

Data Index: 915124 (C4 (Web Text))

Max Activating Token Index: 210

Text #13

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6248. Min Act: -0.0001

Data Index: 27436 (C4 (Web Text))

Max Activating Token Index: 141

Text #14

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6224. Min Act: -0.0001

Data Index: 247207 (C4 (Web Text))

Max Activating Token Index: 301

Text #15

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6626. Min Act: -0.0001

Data Index: 1439876 (Python Code)

Max Activating Token Index: 294

Text #16

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.6179. Min Act: -0.0001

Data Index: 219044 (C4 (Web Text))

Max Activating Token Index: 647

Text #17

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.5621. Min Act: -0.0001

Data Index: 623399 (C4 (Web Text))

Max Activating Token Index: 188

Text #18

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.5774. Min Act: -0.0001

Data Index: 473435 (C4 (Web Text))

Max Activating Token Index: 348

Text #19

Max Range: 0.7427. Min Range: -0.7427

Max Act: 0.5956. Min Act: -0.0001

Data Index: 1352042 (C4 (Web Text))

Max Activating Token Index: 459