Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 417 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0450. Min Act: -0.0006

Data Index: 1544164 (Python Code)

Max Activating Token Index: 252

Text #1

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0459. Min Act: -0.0006

Data Index: 734605 (C4 (Web Text))

Max Activating Token Index: 210

Text #2

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0446. Min Act: -0.0006

Data Index: 455607 (C4 (Web Text))

Max Activating Token Index: 95

Text #3

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0395. Min Act: -0.0005

Data Index: 1619842 (Python Code)

Max Activating Token Index: 26

Text #4

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0385. Min Act: -0.0006

Data Index: 858089 (C4 (Web Text))

Max Activating Token Index: 205

Text #5

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0413. Min Act: -0.0006

Data Index: 1074980 (C4 (Web Text))

Max Activating Token Index: 210

Text #6

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0380. Min Act: -0.0006

Data Index: 1613442 (Python Code)

Max Activating Token Index: 359

Text #7

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0360. Min Act: -0.0006

Data Index: 527782 (C4 (Web Text))

Max Activating Token Index: 384

Text #8

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0359. Min Act: -0.0006

Data Index: 1325748 (C4 (Web Text))

Max Activating Token Index: 156

Text #9

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0354. Min Act: -0.0006

Data Index: 1046206 (C4 (Web Text))

Max Activating Token Index: 807

Text #10

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0349. Min Act: -0.0006

Data Index: 680888 (C4 (Web Text))

Max Activating Token Index: 18

Text #11

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0349. Min Act: -0.0006

Data Index: 439964 (C4 (Web Text))

Max Activating Token Index: 846

Text #12

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0361. Min Act: -0.0006

Data Index: 24951 (C4 (Web Text))

Max Activating Token Index: 834

Text #13

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0344. Min Act: -0.0006

Data Index: 1056026 (C4 (Web Text))

Max Activating Token Index: 104

Text #14

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0359. Min Act: -0.0006

Data Index: 1511853 (Python Code)

Max Activating Token Index: 350

Text #15

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0342. Min Act: -0.0006

Data Index: 868396 (C4 (Web Text))

Max Activating Token Index: 56

Text #16

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0357. Min Act: -0.0006

Data Index: 6724 (C4 (Web Text))

Max Activating Token Index: 483

Text #17

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0313. Min Act: -0.0005

Data Index: 677001 (C4 (Web Text))

Max Activating Token Index: 88

Text #18

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0332. Min Act: -0.0006

Data Index: 1554384 (Python Code)

Max Activating Token Index: 141

Text #19

Max Range: 0.0459. Min Range: -0.0459

Max Act: 0.0331. Min Act: -0.0006

Data Index: 186801 (C4 (Web Text))

Max Activating Token Index: 837