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

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

Neuron 6033 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 4.2317. Min Range: -4.2317

Max Act: 4.2317. Min Act: -0.0001

Data Index: 1006898 (C4 (Web Text))

Max Activating Token Index: 394

Text #1

Max Range: 4.2317. Min Range: -4.2317

Max Act: 3.8860. Min Act: -0.0000

Data Index: 925310 (C4 (Web Text))

Max Activating Token Index: 21

Text #2

Max Range: 4.2317. Min Range: -4.2317

Max Act: 3.6161. Min Act: -0.0000

Data Index: 1262103 (C4 (Web Text))

Max Activating Token Index: 382

Text #3

Max Range: 4.2317. Min Range: -4.2317

Max Act: 3.5095. Min Act: -0.0000

Data Index: 144190 (C4 (Web Text))

Max Activating Token Index: 14

Text #4

Max Range: 4.2317. Min Range: -4.2317

Max Act: 3.3145. Min Act: -0.0000

Data Index: 244422 (C4 (Web Text))

Max Activating Token Index: 63

Text #5

Max Range: 4.2317. Min Range: -4.2317

Max Act: 3.2747. Min Act: -0.0001

Data Index: 918317 (C4 (Web Text))

Max Activating Token Index: 545

Text #6

Max Range: 4.2317. Min Range: -4.2317

Max Act: 3.1982. Min Act: -0.0000

Data Index: 182948 (C4 (Web Text))

Max Activating Token Index: 9

Text #7

Max Range: 4.2317. Min Range: -4.2317

Max Act: 3.2047. Min Act: -0.0000

Data Index: 518185 (C4 (Web Text))

Max Activating Token Index: 5

Text #8

Max Range: 4.2317. Min Range: -4.2317

Max Act: 3.0361. Min Act: -0.0000

Data Index: 1106916 (C4 (Web Text))

Max Activating Token Index: 712

Text #9

Max Range: 4.2317. Min Range: -4.2317

Max Act: 3.0663. Min Act: -0.0001

Data Index: 10769 (C4 (Web Text))

Max Activating Token Index: 482

Text #10

Max Range: 4.2317. Min Range: -4.2317

Max Act: 2.9534. Min Act: -0.0001

Data Index: 677558 (C4 (Web Text))

Max Activating Token Index: 403

Text #11

Max Range: 4.2317. Min Range: -4.2317

Max Act: 2.9461. Min Act: -0.0000

Data Index: 904190 (C4 (Web Text))

Max Activating Token Index: 121

Text #12

Max Range: 4.2317. Min Range: -4.2317

Max Act: 2.9034. Min Act: -0.0000

Data Index: 63623 (C4 (Web Text))

Max Activating Token Index: 462

Text #13

Max Range: 4.2317. Min Range: -4.2317

Max Act: 3.0260. Min Act: -0.0000

Data Index: 202581 (C4 (Web Text))

Max Activating Token Index: 133

Text #14

Max Range: 4.2317. Min Range: -4.2317

Max Act: 2.8502. Min Act: -0.0001

Data Index: 655392 (C4 (Web Text))

Max Activating Token Index: 408

Text #15

Max Range: 4.2317. Min Range: -4.2317

Max Act: 2.6215. Min Act: -0.0001

Data Index: 977985 (C4 (Web Text))

Max Activating Token Index: 424

Text #16

Max Range: 4.2317. Min Range: -4.2317

Max Act: 2.7002. Min Act: -0.0000

Data Index: 420600 (C4 (Web Text))

Max Activating Token Index: 18

Text #17

Max Range: 4.2317. Min Range: -4.2317

Max Act: 2.5586. Min Act: -0.0000

Data Index: 965497 (C4 (Web Text))

Max Activating Token Index: 276

Text #18

Max Range: 4.2317. Min Range: -4.2317

Max Act: 2.6684. Min Act: -0.0000

Data Index: 1353241 (C4 (Web Text))

Max Activating Token Index: 71

Text #19

Max Range: 4.2317. Min Range: -4.2317

Max Act: 2.5849. Min Act: -0.0001

Data Index: 1314596 (C4 (Web Text))

Max Activating Token Index: 580