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

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

Neuron 1638 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: 0.0391. Min Range: -0.0391

Max Act: 0.0391. Min Act: -0.0005

Data Index: 340368 (C4 (Web Text))

Max Activating Token Index: 315

Text #1

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0366. Min Act: -0.0005

Data Index: 1283011 (C4 (Web Text))

Max Activating Token Index: 17

Text #2

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0296. Min Act: -0.0005

Data Index: 294631 (C4 (Web Text))

Max Activating Token Index: 889

Text #3

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0300. Min Act: -0.0005

Data Index: 554287 (C4 (Web Text))

Max Activating Token Index: 204

Text #4

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0269. Min Act: -0.0005

Data Index: 1310801 (C4 (Web Text))

Max Activating Token Index: 805

Text #5

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0250. Min Act: -0.0005

Data Index: 708292 (C4 (Web Text))

Max Activating Token Index: 376

Text #6

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0240. Min Act: -0.0005

Data Index: 989197 (C4 (Web Text))

Max Activating Token Index: 871

Text #7

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0231. Min Act: -0.0005

Data Index: 841222 (C4 (Web Text))

Max Activating Token Index: 245

Text #8

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0237. Min Act: -0.0005

Data Index: 1148345 (C4 (Web Text))

Max Activating Token Index: 978

Text #9

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0229. Min Act: -0.0005

Data Index: 524626 (C4 (Web Text))

Max Activating Token Index: 52

Text #10

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0240. Min Act: -0.0005

Data Index: 865022 (C4 (Web Text))

Max Activating Token Index: 475

Text #11

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0223. Min Act: -0.0005

Data Index: 1075218 (C4 (Web Text))

Max Activating Token Index: 401

Text #12

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0229. Min Act: -0.0005

Data Index: 881707 (C4 (Web Text))

Max Activating Token Index: 360

Text #13

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0221. Min Act: -0.0005

Data Index: 925014 (C4 (Web Text))

Max Activating Token Index: 417

Text #14

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0229. Min Act: -0.0005

Data Index: 1145784 (C4 (Web Text))

Max Activating Token Index: 781

Text #15

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0222. Min Act: -0.0005

Data Index: 1616580 (Python Code)

Max Activating Token Index: 14

Text #16

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0218. Min Act: -0.0005

Data Index: 177506 (C4 (Web Text))

Max Activating Token Index: 333

Text #17

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0203. Min Act: -0.0005

Data Index: 675697 (C4 (Web Text))

Max Activating Token Index: 405

Text #18

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0228. Min Act: -0.0005

Data Index: 239575 (C4 (Web Text))

Max Activating Token Index: 473

Text #19

Max Range: 0.0391. Min Range: -0.0391

Max Act: 0.0209. Min Act: -0.0005

Data Index: 234406 (C4 (Web Text))

Max Activating Token Index: 664