Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 507 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0417. Min Act: -0.0005

Data Index: 221494 (C4 (Web Text))

Max Activating Token Index: 982

Text #1

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0420. Min Act: -0.0005

Data Index: 1028987 (C4 (Web Text))

Max Activating Token Index: 7

Text #2

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0418. Min Act: -0.0005

Data Index: 99110 (C4 (Web Text))

Max Activating Token Index: 348

Text #3

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0404. Min Act: -0.0005

Data Index: 700516 (C4 (Web Text))

Max Activating Token Index: 380

Text #4

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0390. Min Act: -0.0005

Data Index: 614568 (C4 (Web Text))

Max Activating Token Index: 407

Text #5

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0382. Min Act: -0.0005

Data Index: 357438 (C4 (Web Text))

Max Activating Token Index: 243

Text #6

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0380. Min Act: -0.0005

Data Index: 1206001 (C4 (Web Text))

Max Activating Token Index: 362

Text #7

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0384. Min Act: -0.0005

Data Index: 5852 (C4 (Web Text))

Max Activating Token Index: 748

Text #8

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0378. Min Act: -0.0005

Data Index: 492610 (C4 (Web Text))

Max Activating Token Index: 309

Text #9

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0368. Min Act: -0.0005

Data Index: 1224701 (C4 (Web Text))

Max Activating Token Index: 437

Text #10

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0373. Min Act: -0.0005

Data Index: 968612 (C4 (Web Text))

Max Activating Token Index: 392

Text #11

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0377. Min Act: -0.0005

Data Index: 392158 (C4 (Web Text))

Max Activating Token Index: 866

Text #12

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0362. Min Act: -0.0005

Data Index: 47399 (C4 (Web Text))

Max Activating Token Index: 442

Text #13

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0376. Min Act: -0.0005

Data Index: 1604745 (Python Code)

Max Activating Token Index: 386

Text #14

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0357. Min Act: -0.0005

Data Index: 35681 (C4 (Web Text))

Max Activating Token Index: 834

Text #15

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0352. Min Act: -0.0005

Data Index: 1331481 (C4 (Web Text))

Max Activating Token Index: 937

Text #16

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0352. Min Act: -0.0005

Data Index: 904971 (C4 (Web Text))

Max Activating Token Index: 310

Text #17

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0344. Min Act: -0.0005

Data Index: 288567 (C4 (Web Text))

Max Activating Token Index: 283

Text #18

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0371. Min Act: -0.0005

Data Index: 725229 (C4 (Web Text))

Max Activating Token Index: 686

Text #19

Max Range: 0.0420. Min Range: -0.0420

Max Act: 0.0347. Min Act: -0.0005

Data Index: 185855 (C4 (Web Text))

Max Activating Token Index: 23