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

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

Neuron 2050 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0580. Min Act: -0.0003

Data Index: 182993 (C4 (Web Text))

Max Activating Token Index: 165

Text #1

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0467. Min Act: -0.0003

Data Index: 973434 (C4 (Web Text))

Max Activating Token Index: 365

Text #2

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0493. Min Act: -0.0003

Data Index: 642031 (C4 (Web Text))

Max Activating Token Index: 99

Text #3

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0432. Min Act: -0.0002

Data Index: 817749 (C4 (Web Text))

Max Activating Token Index: 888

Text #4

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0456. Min Act: -0.0003

Data Index: 1174731 (C4 (Web Text))

Max Activating Token Index: 54

Text #5

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0427. Min Act: -0.0002

Data Index: 784081 (C4 (Web Text))

Max Activating Token Index: 502

Text #6

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0448. Min Act: -0.0003

Data Index: 599443 (C4 (Web Text))

Max Activating Token Index: 1023

Text #7

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0326. Min Act: -0.0003

Data Index: 1321944 (C4 (Web Text))

Max Activating Token Index: 280

Text #8

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0498. Min Act: -0.0003

Data Index: 1639850 (Python Code)

Max Activating Token Index: 415

Text #9

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0452. Min Act: -0.0003

Data Index: 1545923 (Python Code)

Max Activating Token Index: 495

Text #10

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0435. Min Act: -0.0003

Data Index: 1193751 (C4 (Web Text))

Max Activating Token Index: 562

Text #11

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0495. Min Act: -0.0003

Data Index: 668648 (C4 (Web Text))

Max Activating Token Index: 157

Text #12

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0344. Min Act: -0.0003

Data Index: 565035 (C4 (Web Text))

Max Activating Token Index: 634

Text #13

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0434. Min Act: -0.0003

Data Index: 1277405 (C4 (Web Text))

Max Activating Token Index: 772

Text #14

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0506. Min Act: -0.0003

Data Index: 213218 (C4 (Web Text))

Max Activating Token Index: 232

Text #15

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0447. Min Act: -0.0002

Data Index: 1629673 (Python Code)

Max Activating Token Index: 440

Text #16

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0409. Min Act: -0.0002

Data Index: 650577 (C4 (Web Text))

Max Activating Token Index: 593

Text #17

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0411. Min Act: -0.0003

Data Index: 428803 (C4 (Web Text))

Max Activating Token Index: 54

Text #18

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0395. Min Act: -0.0003

Data Index: 1220386 (C4 (Web Text))

Max Activating Token Index: 655

Text #19

Max Range: 0.0580. Min Range: -0.0580

Max Act: 0.0340. Min Act: -0.0003

Data Index: 192285 (C4 (Web Text))

Max Activating Token Index: 323