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

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

Neuron 2655 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.6419. Min Range: -3.6419

Max Act: 3.5799. Min Act: -0.0001

Data Index: 1488033 (Python Code)

Max Activating Token Index: 323

Text #1

Max Range: 3.6419. Min Range: -3.6419

Max Act: 3.6419. Min Act: -0.0000

Data Index: 1315032 (C4 (Web Text))

Max Activating Token Index: 210

Text #2

Max Range: 3.6419. Min Range: -3.6419

Max Act: 3.6109. Min Act: -0.0000

Data Index: 830135 (C4 (Web Text))

Max Activating Token Index: 160

Text #3

Max Range: 3.6419. Min Range: -3.6419

Max Act: 3.0810. Min Act: -0.0000

Data Index: 629004 (C4 (Web Text))

Max Activating Token Index: 328

Text #4

Max Range: 3.6419. Min Range: -3.6419

Max Act: 3.4439. Min Act: -0.0000

Data Index: 1343448 (C4 (Web Text))

Max Activating Token Index: 126

Text #5

Max Range: 3.6419. Min Range: -3.6419

Max Act: 3.4395. Min Act: -0.0000

Data Index: 877960 (C4 (Web Text))

Max Activating Token Index: 869

Text #6

Max Range: 3.6419. Min Range: -3.6419

Max Act: 2.9711. Min Act: -0.0001

Data Index: 1579528 (Python Code)

Max Activating Token Index: 976

Text #7

Max Range: 3.6419. Min Range: -3.6419

Max Act: 3.0913. Min Act: -0.0001

Data Index: 1550840 (Python Code)

Max Activating Token Index: 893

Text #8

Max Range: 3.6419. Min Range: -3.6419

Max Act: 2.8059. Min Act: -0.0001

Data Index: 1621240 (Python Code)

Max Activating Token Index: 230

Text #9

Max Range: 3.6419. Min Range: -3.6419

Max Act: 2.8965. Min Act: -0.0001

Data Index: 1646973 (Python Code)

Max Activating Token Index: 244

Text #10

Max Range: 3.6419. Min Range: -3.6419

Max Act: 3.0099. Min Act: -0.0001

Data Index: 868927 (C4 (Web Text))

Max Activating Token Index: 76

Text #11

Max Range: 3.6419. Min Range: -3.6419

Max Act: 2.8667. Min Act: -0.0000

Data Index: 157496 (C4 (Web Text))

Max Activating Token Index: 294

Text #12

Max Range: 3.6419. Min Range: -3.6419

Max Act: 3.1005. Min Act: -0.0001

Data Index: 466868 (C4 (Web Text))

Max Activating Token Index: 315

Text #13

Max Range: 3.6419. Min Range: -3.6419

Max Act: 2.9648. Min Act: -0.0000

Data Index: 1600616 (Python Code)

Max Activating Token Index: 391

Text #14

Max Range: 3.6419. Min Range: -3.6419

Max Act: 3.1042. Min Act: -0.0000

Data Index: 1609704 (Python Code)

Max Activating Token Index: 269

Text #15

Max Range: 3.6419. Min Range: -3.6419

Max Act: 2.7273. Min Act: -0.0001

Data Index: 297983 (C4 (Web Text))

Max Activating Token Index: 979

Text #16

Max Range: 3.6419. Min Range: -3.6419

Max Act: 2.7363. Min Act: -0.0000

Data Index: 897652 (C4 (Web Text))

Max Activating Token Index: 888

Text #17

Max Range: 3.6419. Min Range: -3.6419

Max Act: 3.0044. Min Act: -0.0001

Data Index: 1638354 (Python Code)

Max Activating Token Index: 237

Text #18

Max Range: 3.6419. Min Range: -3.6419

Max Act: 2.8452. Min Act: -0.0000

Data Index: 1489962 (Python Code)

Max Activating Token Index: 99

Text #19

Max Range: 3.6419. Min Range: -3.6419

Max Act: 2.7868. Min Act: -0.0000

Data Index: 391953 (C4 (Web Text))

Max Activating Token Index: 90