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

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

Neuron 3485 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.5090. Min Act: -0.0000

Data Index: 991696 (C4 (Web Text))

Max Activating Token Index: 474

Text #1

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.4803. Min Act: -0.0000

Data Index: 1533083 (Python Code)

Max Activating Token Index: 332

Text #2

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.4467. Min Act: -0.0000

Data Index: 1582906 (Python Code)

Max Activating Token Index: 301

Text #3

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.4384. Min Act: -0.0000

Data Index: 311914 (C4 (Web Text))

Max Activating Token Index: 373

Text #4

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.4240. Min Act: -0.0000

Data Index: 1382170 (Python Code)

Max Activating Token Index: 731

Text #5

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.3712. Min Act: -0.0000

Data Index: 321531 (C4 (Web Text))

Max Activating Token Index: 127

Text #6

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.4237. Min Act: -0.0000

Data Index: 475643 (C4 (Web Text))

Max Activating Token Index: 70

Text #7

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.4200. Min Act: -0.0000

Data Index: 774162 (C4 (Web Text))

Max Activating Token Index: 971

Text #8

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.4048. Min Act: -0.0000

Data Index: 1572292 (Python Code)

Max Activating Token Index: 297

Text #9

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.4107. Min Act: -0.0000

Data Index: 1290168 (C4 (Web Text))

Max Activating Token Index: 193

Text #10

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.3768. Min Act: -0.0000

Data Index: 641165 (C4 (Web Text))

Max Activating Token Index: 155

Text #11

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.3733. Min Act: -0.0000

Data Index: 818073 (C4 (Web Text))

Max Activating Token Index: 157

Text #12

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.4056. Min Act: -0.0000

Data Index: 1186103 (C4 (Web Text))

Max Activating Token Index: 73

Text #13

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.3566. Min Act: -0.0000

Data Index: 255128 (C4 (Web Text))

Max Activating Token Index: 320

Text #14

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.3676. Min Act: -0.0000

Data Index: 542170 (C4 (Web Text))

Max Activating Token Index: 23

Text #15

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.3950. Min Act: -0.0000

Data Index: 1397288 (Python Code)

Max Activating Token Index: 613

Text #16

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.3608. Min Act: -0.0000

Data Index: 1567882 (Python Code)

Max Activating Token Index: 720

Text #17

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.3873. Min Act: -0.0000

Data Index: 1239806 (C4 (Web Text))

Max Activating Token Index: 117

Text #18

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.3610. Min Act: -0.0000

Data Index: 626450 (C4 (Web Text))

Max Activating Token Index: 81

Text #19

Max Range: 0.5090. Min Range: -0.5090

Max Act: 0.3751. Min Act: -0.0000

Data Index: 637382 (C4 (Web Text))

Max Activating Token Index: 411