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

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

Neuron 2020 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 6.1272. Min Range: -6.1272

Max Act: 6.1272. Min Act: -0.0000

Data Index: 1127518 (C4 (Web Text))

Max Activating Token Index: 212

Text #1

Max Range: 6.1272. Min Range: -6.1272

Max Act: 4.8898. Min Act: -0.0001

Data Index: 1524104 (Python Code)

Max Activating Token Index: 143

Text #2

Max Range: 6.1272. Min Range: -6.1272

Max Act: 5.4981. Min Act: -0.0001

Data Index: 1404633 (Python Code)

Max Activating Token Index: 608

Text #3

Max Range: 6.1272. Min Range: -6.1272

Max Act: 5.7839. Min Act: 0.0001

Data Index: 1561648 (Python Code)

Max Activating Token Index: 309

Text #4

Max Range: 6.1272. Min Range: -6.1272

Max Act: 5.0268. Min Act: -0.0001

Data Index: 1382131 (Python Code)

Max Activating Token Index: 14

Text #5

Max Range: 6.1272. Min Range: -6.1272

Max Act: 4.4084. Min Act: -0.0001

Data Index: 1635466 (Python Code)

Max Activating Token Index: 629

Text #6

Max Range: 6.1272. Min Range: -6.1272

Max Act: 4.7841. Min Act: 0.0001

Data Index: 1525476 (Python Code)

Max Activating Token Index: 42

Text #7

Max Range: 6.1272. Min Range: -6.1272

Max Act: 4.2771. Min Act: -0.0001

Data Index: 1623860 (Python Code)

Max Activating Token Index: 875

Text #8

Max Range: 6.1272. Min Range: -6.1272

Max Act: 4.2027. Min Act: -0.0000

Data Index: 163318 (C4 (Web Text))

Max Activating Token Index: 519

Text #9

Max Range: 6.1272. Min Range: -6.1272

Max Act: 4.3422. Min Act: -0.0000

Data Index: 1456271 (Python Code)

Max Activating Token Index: 189

Text #10

Max Range: 6.1272. Min Range: -6.1272

Max Act: 4.1713. Min Act: -0.0001

Data Index: 1646404 (Python Code)

Max Activating Token Index: 114

Text #11

Max Range: 6.1272. Min Range: -6.1272

Max Act: 4.0534. Min Act: -0.0000

Data Index: 1143120 (C4 (Web Text))

Max Activating Token Index: 119

Text #12

Max Range: 6.1272. Min Range: -6.1272

Max Act: 4.2271. Min Act: -0.0001

Data Index: 1075235 (C4 (Web Text))

Max Activating Token Index: 151

Text #13

Max Range: 6.1272. Min Range: -6.1272

Max Act: 4.0156. Min Act: -0.0001

Data Index: 1489183 (Python Code)

Max Activating Token Index: 487

Text #14

Max Range: 6.1272. Min Range: -6.1272

Max Act: 3.6745. Min Act: -0.0001

Data Index: 1404016 (Python Code)

Max Activating Token Index: 61

Text #15

Max Range: 6.1272. Min Range: -6.1272

Max Act: 3.6256. Min Act: -0.0000

Data Index: 1272077 (C4 (Web Text))

Max Activating Token Index: 74

Text #16

Max Range: 6.1272. Min Range: -6.1272

Max Act: 4.0806. Min Act: 0.0000

Data Index: 1385973 (Python Code)

Max Activating Token Index: 828

Text #17

Max Range: 6.1272. Min Range: -6.1272

Max Act: 3.8293. Min Act: -0.0001

Data Index: 1606018 (Python Code)

Max Activating Token Index: 58

Text #18

Max Range: 6.1272. Min Range: -6.1272

Max Act: 3.6699. Min Act: -0.0000

Data Index: 1160414 (C4 (Web Text))

Max Activating Token Index: 459

Text #19

Max Range: 6.1272. Min Range: -6.1272

Max Act: 2.9563. Min Act: -0.0001

Data Index: 1433204 (Python Code)

Max Activating Token Index: 804