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

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

Neuron 1060 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: 0.0305. Min Range: -0.0305

Max Act: 0.0304. Min Act: -0.0003

Data Index: 592114 (C4 (Web Text))

Max Activating Token Index: 876

Text #1

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0305. Min Act: -0.0003

Data Index: 1109449 (C4 (Web Text))

Max Activating Token Index: 183

Text #2

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0295. Min Act: -0.0003

Data Index: 604202 (C4 (Web Text))

Max Activating Token Index: 669

Text #3

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0287. Min Act: -0.0003

Data Index: 701743 (C4 (Web Text))

Max Activating Token Index: 87

Text #4

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0267. Min Act: -0.0003

Data Index: 89692 (C4 (Web Text))

Max Activating Token Index: 877

Text #5

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0267. Min Act: -0.0003

Data Index: 1204632 (C4 (Web Text))

Max Activating Token Index: 635

Text #6

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0263. Min Act: -0.0003

Data Index: 555744 (C4 (Web Text))

Max Activating Token Index: 98

Text #7

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0262. Min Act: -0.0003

Data Index: 721848 (C4 (Web Text))

Max Activating Token Index: 437

Text #8

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0268. Min Act: -0.0003

Data Index: 424148 (C4 (Web Text))

Max Activating Token Index: 147

Text #9

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0258. Min Act: -0.0003

Data Index: 369125 (C4 (Web Text))

Max Activating Token Index: 341

Text #10

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0265. Min Act: -0.0003

Data Index: 1633584 (Python Code)

Max Activating Token Index: 743

Text #11

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0255. Min Act: -0.0003

Data Index: 35608 (C4 (Web Text))

Max Activating Token Index: 342

Text #12

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0260. Min Act: -0.0003

Data Index: 1589875 (Python Code)

Max Activating Token Index: 983

Text #13

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0261. Min Act: -0.0003

Data Index: 500630 (C4 (Web Text))

Max Activating Token Index: 11

Text #14

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0258. Min Act: -0.0003

Data Index: 111032 (C4 (Web Text))

Max Activating Token Index: 664

Text #15

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0247. Min Act: -0.0004

Data Index: 57488 (C4 (Web Text))

Max Activating Token Index: 1023

Text #16

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0256. Min Act: -0.0003

Data Index: 1588742 (Python Code)

Max Activating Token Index: 558

Text #17

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0255. Min Act: -0.0003

Data Index: 1047159 (C4 (Web Text))

Max Activating Token Index: 553

Text #18

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0254. Min Act: -0.0003

Data Index: 588695 (C4 (Web Text))

Max Activating Token Index: 455

Text #19

Max Range: 0.0305. Min Range: -0.0305

Max Act: 0.0246. Min Act: -0.0002

Data Index: 337861 (C4 (Web Text))

Max Activating Token Index: 934