Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 443 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 6.6235. Min Range: -6.6235

Max Act: 6.6235. Min Act: -0.0000

Data Index: 885509 (C4 (Web Text))

Max Activating Token Index: 770

Text #1

Max Range: 6.6235. Min Range: -6.6235

Max Act: 6.1236. Min Act: -0.0000

Data Index: 1138144 (C4 (Web Text))

Max Activating Token Index: 442

Text #2

Max Range: 6.6235. Min Range: -6.6235

Max Act: 5.8072. Min Act: -0.0000

Data Index: 193800 (C4 (Web Text))

Max Activating Token Index: 377

Text #3

Max Range: 6.6235. Min Range: -6.6235

Max Act: 5.3682. Min Act: -0.0000

Data Index: 1607981 (Python Code)

Max Activating Token Index: 115

Text #4

Max Range: 6.6235. Min Range: -6.6235

Max Act: 5.5542. Min Act: -0.0000

Data Index: 1420962 (Python Code)

Max Activating Token Index: 289

Text #5

Max Range: 6.6235. Min Range: -6.6235

Max Act: 5.5513. Min Act: -0.0000

Data Index: 586366 (C4 (Web Text))

Max Activating Token Index: 873

Text #6

Max Range: 6.6235. Min Range: -6.6235

Max Act: 5.1893. Min Act: -0.0000

Data Index: 1565234 (Python Code)

Max Activating Token Index: 357

Text #7

Max Range: 6.6235. Min Range: -6.6235

Max Act: 5.2807. Min Act: -0.0000

Data Index: 938326 (C4 (Web Text))

Max Activating Token Index: 858

Text #8

Max Range: 6.6235. Min Range: -6.6235

Max Act: 5.0907. Min Act: -0.0000

Data Index: 752134 (C4 (Web Text))

Max Activating Token Index: 839

Text #9

Max Range: 6.6235. Min Range: -6.6235

Max Act: 5.4525. Min Act: -0.0000

Data Index: 283240 (C4 (Web Text))

Max Activating Token Index: 760

Text #10

Max Range: 6.6235. Min Range: -6.6235

Max Act: 5.4228. Min Act: -0.0000

Data Index: 1069089 (C4 (Web Text))

Max Activating Token Index: 660

Text #11

Max Range: 6.6235. Min Range: -6.6235

Max Act: 5.0665. Min Act: -0.0000

Data Index: 1567664 (Python Code)

Max Activating Token Index: 481

Text #12

Max Range: 6.6235. Min Range: -6.6235

Max Act: 4.9230. Min Act: -0.0000

Data Index: 910915 (C4 (Web Text))

Max Activating Token Index: 573

Text #13

Max Range: 6.6235. Min Range: -6.6235

Max Act: 4.7294. Min Act: -0.0000

Data Index: 1177723 (C4 (Web Text))

Max Activating Token Index: 210

Text #14

Max Range: 6.6235. Min Range: -6.6235

Max Act: 5.0719. Min Act: -0.0000

Data Index: 1572681 (Python Code)

Max Activating Token Index: 245

Text #15

Max Range: 6.6235. Min Range: -6.6235

Max Act: 4.7954. Min Act: -0.0000

Data Index: 1025341 (C4 (Web Text))

Max Activating Token Index: 938

Text #16

Max Range: 6.6235. Min Range: -6.6235

Max Act: 4.8267. Min Act: -0.0000

Data Index: 1492787 (Python Code)

Max Activating Token Index: 965

Text #17

Max Range: 6.6235. Min Range: -6.6235

Max Act: 5.0128. Min Act: -0.0000

Data Index: 439859 (C4 (Web Text))

Max Activating Token Index: 364

Text #18

Max Range: 6.6235. Min Range: -6.6235

Max Act: 4.5804. Min Act: -0.0000

Data Index: 910958 (C4 (Web Text))

Max Activating Token Index: 186

Text #19

Max Range: 6.6235. Min Range: -6.6235

Max Act: 4.7426. Min Act: -0.0000

Data Index: 1361487 (Python Code)

Max Activating Token Index: 799