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

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

Neuron 1814 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0431. Min Act: -0.0003

Data Index: 1605044 (Python Code)

Max Activating Token Index: 486

Text #1

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0414. Min Act: -0.0003

Data Index: 774331 (C4 (Web Text))

Max Activating Token Index: 1023

Text #2

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0394. Min Act: -0.0003

Data Index: 1619339 (Python Code)

Max Activating Token Index: 771

Text #3

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0388. Min Act: -0.0002

Data Index: 615018 (C4 (Web Text))

Max Activating Token Index: 856

Text #4

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0369. Min Act: -0.0003

Data Index: 1065145 (C4 (Web Text))

Max Activating Token Index: 938

Text #5

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0372. Min Act: -0.0003

Data Index: 250583 (C4 (Web Text))

Max Activating Token Index: 766

Text #6

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0365. Min Act: -0.0003

Data Index: 1182391 (C4 (Web Text))

Max Activating Token Index: 370

Text #7

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0365. Min Act: -0.0002

Data Index: 1335213 (C4 (Web Text))

Max Activating Token Index: 954

Text #8

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0343. Min Act: -0.0003

Data Index: 722787 (C4 (Web Text))

Max Activating Token Index: 954

Text #9

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0360. Min Act: -0.0002

Data Index: 422260 (C4 (Web Text))

Max Activating Token Index: 268

Text #10

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0362. Min Act: -0.0004

Data Index: 827348 (C4 (Web Text))

Max Activating Token Index: 318

Text #11

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0349. Min Act: -0.0002

Data Index: 328000 (C4 (Web Text))

Max Activating Token Index: 242

Text #12

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0347. Min Act: -0.0003

Data Index: 615119 (C4 (Web Text))

Max Activating Token Index: 436

Text #13

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0354. Min Act: -0.0003

Data Index: 1287800 (C4 (Web Text))

Max Activating Token Index: 744

Text #14

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0359. Min Act: -0.0002

Data Index: 468117 (C4 (Web Text))

Max Activating Token Index: 54

Text #15

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0349. Min Act: -0.0003

Data Index: 1653620 (Python Code)

Max Activating Token Index: 360

Text #16

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0353. Min Act: -0.0003

Data Index: 168178 (C4 (Web Text))

Max Activating Token Index: 746

Text #17

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0348. Min Act: -0.0003

Data Index: 887657 (C4 (Web Text))

Max Activating Token Index: 553

Text #18

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0345. Min Act: -0.0002

Data Index: 1557866 (Python Code)

Max Activating Token Index: 600

Text #19

Max Range: 0.0431. Min Range: -0.0431

Max Act: 0.0338. Min Act: -0.0002

Data Index: 543623 (C4 (Web Text))

Max Activating Token Index: 541