Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 4465 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 2.3824. Min Range: -2.3824

Max Act: 2.3824. Min Act: -0.0000

Data Index: 1137376 (C4 (Web Text))

Max Activating Token Index: 148

Text #1

Max Range: 2.3824. Min Range: -2.3824

Max Act: 2.0343. Min Act: -0.0001

Data Index: 739395 (C4 (Web Text))

Max Activating Token Index: 474

Text #2

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.7746. Min Act: -0.0000

Data Index: 1003068 (C4 (Web Text))

Max Activating Token Index: 12

Text #3

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.6263. Min Act: -0.0000

Data Index: 459497 (C4 (Web Text))

Max Activating Token Index: 7

Text #4

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.5329. Min Act: -0.0000

Data Index: 1539053 (Python Code)

Max Activating Token Index: 179

Text #5

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.6617. Min Act: -0.0000

Data Index: 518439 (C4 (Web Text))

Max Activating Token Index: 204

Text #6

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.4079. Min Act: -0.0000

Data Index: 338704 (C4 (Web Text))

Max Activating Token Index: 289

Text #7

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.5128. Min Act: -0.0001

Data Index: 561342 (C4 (Web Text))

Max Activating Token Index: 33

Text #8

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.7467. Min Act: -0.0000

Data Index: 1494293 (Python Code)

Max Activating Token Index: 91

Text #9

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.5401. Min Act: -0.0000

Data Index: 1568797 (Python Code)

Max Activating Token Index: 67

Text #10

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.6331. Min Act: -0.0001

Data Index: 462968 (C4 (Web Text))

Max Activating Token Index: 8

Text #11

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.4322. Min Act: -0.0001

Data Index: 1562002 (Python Code)

Max Activating Token Index: 21

Text #12

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.3122. Min Act: -0.0000

Data Index: 622625 (C4 (Web Text))

Max Activating Token Index: 856

Text #13

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.4751. Min Act: -0.0000

Data Index: 1528419 (Python Code)

Max Activating Token Index: 509

Text #14

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.1023. Min Act: -0.0001

Data Index: 680291 (C4 (Web Text))

Max Activating Token Index: 667

Text #15

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.4261. Min Act: -0.0001

Data Index: 578556 (C4 (Web Text))

Max Activating Token Index: 67

Text #16

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.2717. Min Act: -0.0000

Data Index: 499317 (C4 (Web Text))

Max Activating Token Index: 206

Text #17

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.5036. Min Act: -0.0001

Data Index: 752751 (C4 (Web Text))

Max Activating Token Index: 516

Text #18

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.3741. Min Act: -0.0000

Data Index: 564042 (C4 (Web Text))

Max Activating Token Index: 55

Text #19

Max Range: 2.3824. Min Range: -2.3824

Max Act: 1.3316. Min Act: -0.0001

Data Index: 1553516 (Python Code)

Max Activating Token Index: 13