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

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

Neuron 4771 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.1423. Min Range: -3.1423

Max Act: 3.1423. Min Act: -0.0000

Data Index: 1253061 (C4 (Web Text))

Max Activating Token Index: 629

Text #1

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.8424. Min Act: -0.0000

Data Index: 964887 (C4 (Web Text))

Max Activating Token Index: 651

Text #2

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.8397. Min Act: -0.0000

Data Index: 170791 (C4 (Web Text))

Max Activating Token Index: 843

Text #3

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.5799. Min Act: -0.0000

Data Index: 686584 (C4 (Web Text))

Max Activating Token Index: 309

Text #4

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.6777. Min Act: -0.0000

Data Index: 1497215 (Python Code)

Max Activating Token Index: 996

Text #5

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.4941. Min Act: -0.0000

Data Index: 1333367 (C4 (Web Text))

Max Activating Token Index: 845

Text #6

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.7505. Min Act: -0.0000

Data Index: 20330 (C4 (Web Text))

Max Activating Token Index: 255

Text #7

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.5790. Min Act: -0.0000

Data Index: 1141271 (C4 (Web Text))

Max Activating Token Index: 96

Text #8

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.4560. Min Act: -0.0000

Data Index: 1586363 (Python Code)

Max Activating Token Index: 521

Text #9

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.4455. Min Act: -0.0000

Data Index: 1029688 (C4 (Web Text))

Max Activating Token Index: 261

Text #10

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.6674. Min Act: -0.0000

Data Index: 1187051 (C4 (Web Text))

Max Activating Token Index: 143

Text #11

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.3322. Min Act: -0.0000

Data Index: 1187165 (C4 (Web Text))

Max Activating Token Index: 526

Text #12

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.4420. Min Act: -0.0000

Data Index: 1378127 (Python Code)

Max Activating Token Index: 803

Text #13

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.4169. Min Act: -0.0000

Data Index: 174089 (C4 (Web Text))

Max Activating Token Index: 233

Text #14

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.4190. Min Act: -0.0000

Data Index: 560414 (C4 (Web Text))

Max Activating Token Index: 563

Text #15

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.2709. Min Act: -0.0000

Data Index: 842868 (C4 (Web Text))

Max Activating Token Index: 952

Text #16

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.2702. Min Act: -0.0000

Data Index: 1325543 (C4 (Web Text))

Max Activating Token Index: 748

Text #17

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.3800. Min Act: -0.0000

Data Index: 440611 (C4 (Web Text))

Max Activating Token Index: 854

Text #18

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.2126. Min Act: -0.0000

Data Index: 1043916 (C4 (Web Text))

Max Activating Token Index: 534

Text #19

Max Range: 3.1423. Min Range: -3.1423

Max Act: 2.2124. Min Act: -0.0000

Data Index: 1295408 (C4 (Web Text))

Max Activating Token Index: 987