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

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

Neuron 2105 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0795. Min Act: -0.0001

Data Index: 277926 (C4 (Web Text))

Max Activating Token Index: 137

Text #1

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0738. Min Act: -0.0001

Data Index: 263283 (C4 (Web Text))

Max Activating Token Index: 71

Text #2

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0541. Min Act: -0.0001

Data Index: 198505 (C4 (Web Text))

Max Activating Token Index: 241

Text #3

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0596. Min Act: -0.0001

Data Index: 432447 (C4 (Web Text))

Max Activating Token Index: 520

Text #4

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0584. Min Act: -0.0001

Data Index: 278514 (C4 (Web Text))

Max Activating Token Index: 561

Text #5

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0617. Min Act: -0.0001

Data Index: 409905 (C4 (Web Text))

Max Activating Token Index: 33

Text #6

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0630. Min Act: -0.0001

Data Index: 403322 (C4 (Web Text))

Max Activating Token Index: 129

Text #7

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0536. Min Act: -0.0001

Data Index: 1030818 (C4 (Web Text))

Max Activating Token Index: 614

Text #8

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0527. Min Act: -0.0001

Data Index: 123127 (C4 (Web Text))

Max Activating Token Index: 451

Text #9

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0541. Min Act: -0.0001

Data Index: 1294009 (C4 (Web Text))

Max Activating Token Index: 30

Text #10

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0536. Min Act: -0.0001

Data Index: 1615165 (Python Code)

Max Activating Token Index: 268

Text #11

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0586. Min Act: -0.0001

Data Index: 403049 (C4 (Web Text))

Max Activating Token Index: 333

Text #12

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0505. Min Act: -0.0001

Data Index: 1420626 (Python Code)

Max Activating Token Index: 339

Text #13

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0524. Min Act: -0.0001

Data Index: 1656181 (Python Code)

Max Activating Token Index: 595

Text #14

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0492. Min Act: -0.0001

Data Index: 860283 (C4 (Web Text))

Max Activating Token Index: 874

Text #15

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0523. Min Act: -0.0001

Data Index: 432041 (C4 (Web Text))

Max Activating Token Index: 317

Text #16

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0534. Min Act: -0.0001

Data Index: 812985 (C4 (Web Text))

Max Activating Token Index: 444

Text #17

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0549. Min Act: -0.0001

Data Index: 745075 (C4 (Web Text))

Max Activating Token Index: 911

Text #18

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0486. Min Act: -0.0001

Data Index: 380353 (C4 (Web Text))

Max Activating Token Index: 14

Text #19

Max Range: 0.0795. Min Range: -0.0795

Max Act: 0.0488. Min Act: -0.0001

Data Index: 607868 (C4 (Web Text))

Max Activating Token Index: 6