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

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

Neuron 1789 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: 0.0384. Min Range: -0.0384

Max Act: 0.0376. Min Act: -0.0006

Data Index: 13504 (C4 (Web Text))

Max Activating Token Index: 354

Text #1

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0384. Min Act: -0.0006

Data Index: 743779 (C4 (Web Text))

Max Activating Token Index: 23

Text #2

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0378. Min Act: -0.0006

Data Index: 201209 (C4 (Web Text))

Max Activating Token Index: 139

Text #3

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0380. Min Act: -0.0005

Data Index: 834374 (C4 (Web Text))

Max Activating Token Index: 425

Text #4

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0378. Min Act: -0.0006

Data Index: 1295136 (C4 (Web Text))

Max Activating Token Index: 365

Text #5

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0370. Min Act: -0.0005

Data Index: 1057711 (C4 (Web Text))

Max Activating Token Index: 256

Text #6

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0370. Min Act: -0.0006

Data Index: 393087 (C4 (Web Text))

Max Activating Token Index: 947

Text #7

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0360. Min Act: -0.0006

Data Index: 373006 (C4 (Web Text))

Max Activating Token Index: 718

Text #8

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0351. Min Act: -0.0006

Data Index: 90935 (C4 (Web Text))

Max Activating Token Index: 286

Text #9

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0350. Min Act: -0.0006

Data Index: 829343 (C4 (Web Text))

Max Activating Token Index: 974

Text #10

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0350. Min Act: -0.0005

Data Index: 540077 (C4 (Web Text))

Max Activating Token Index: 135

Text #11

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0355. Min Act: -0.0005

Data Index: 1262579 (C4 (Web Text))

Max Activating Token Index: 287

Text #12

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0348. Min Act: -0.0006

Data Index: 1102946 (C4 (Web Text))

Max Activating Token Index: 666

Text #13

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0341. Min Act: -0.0006

Data Index: 1208591 (C4 (Web Text))

Max Activating Token Index: 386

Text #14

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0352. Min Act: -0.0005

Data Index: 1492341 (Python Code)

Max Activating Token Index: 700

Text #15

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0367. Min Act: -0.0005

Data Index: 794358 (C4 (Web Text))

Max Activating Token Index: 142

Text #16

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0343. Min Act: -0.0006

Data Index: 1224933 (C4 (Web Text))

Max Activating Token Index: 329

Text #17

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0336. Min Act: -0.0005

Data Index: 535085 (C4 (Web Text))

Max Activating Token Index: 523

Text #18

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0347. Min Act: -0.0006

Data Index: 106517 (C4 (Web Text))

Max Activating Token Index: 386

Text #19

Max Range: 0.0384. Min Range: -0.0384

Max Act: 0.0355. Min Act: -0.0006

Data Index: 1312569 (C4 (Web Text))

Max Activating Token Index: 772