Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 568 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.6309. Min Act: -0.0001

Data Index: 1373725 (Python Code)

Max Activating Token Index: 374

Text #1

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.6188. Min Act: -0.0001

Data Index: 8092 (C4 (Web Text))

Max Activating Token Index: 281

Text #2

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.6502. Min Act: -0.0001

Data Index: 1079804 (C4 (Web Text))

Max Activating Token Index: 564

Text #3

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.6043. Min Act: -0.0000

Data Index: 790540 (C4 (Web Text))

Max Activating Token Index: 222

Text #4

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.6226. Min Act: -0.0000

Data Index: 447488 (C4 (Web Text))

Max Activating Token Index: 467

Text #5

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.6211. Min Act: -0.0001

Data Index: 1335325 (C4 (Web Text))

Max Activating Token Index: 869

Text #6

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5795. Min Act: -0.0000

Data Index: 138637 (C4 (Web Text))

Max Activating Token Index: 780

Text #7

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5986. Min Act: -0.0001

Data Index: 1576367 (Python Code)

Max Activating Token Index: 527

Text #8

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5955. Min Act: -0.0001

Data Index: 134565 (C4 (Web Text))

Max Activating Token Index: 171

Text #9

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5898. Min Act: -0.0001

Data Index: 1306902 (C4 (Web Text))

Max Activating Token Index: 626

Text #10

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5854. Min Act: -0.0000

Data Index: 167787 (C4 (Web Text))

Max Activating Token Index: 170

Text #11

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5634. Min Act: -0.0001

Data Index: 392215 (C4 (Web Text))

Max Activating Token Index: 294

Text #12

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5814. Min Act: -0.0000

Data Index: 1044518 (C4 (Web Text))

Max Activating Token Index: 9

Text #13

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5599. Min Act: -0.0001

Data Index: 281868 (C4 (Web Text))

Max Activating Token Index: 293

Text #14

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5772. Min Act: -0.0000

Data Index: 20678 (C4 (Web Text))

Max Activating Token Index: 223

Text #15

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5718. Min Act: -0.0001

Data Index: 270074 (C4 (Web Text))

Max Activating Token Index: 335

Text #16

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5528. Min Act: -0.0001

Data Index: 572241 (C4 (Web Text))

Max Activating Token Index: 119

Text #17

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5707. Min Act: -0.0001

Data Index: 1019272 (C4 (Web Text))

Max Activating Token Index: 605

Text #18

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5706. Min Act: -0.0001

Data Index: 355031 (C4 (Web Text))

Max Activating Token Index: 521

Text #19

Max Range: 0.6502. Min Range: -0.6502

Max Act: 0.5662. Min Act: -0.0001

Data Index: 1011584 (C4 (Web Text))

Max Activating Token Index: 972