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

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

Neuron 2820 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: 1.6074. Min Range: -1.6074

Max Act: 1.6074. Min Act: -0.0001

Data Index: 755763 (C4 (Web Text))

Max Activating Token Index: 755

Text #1

Max Range: 1.6074. Min Range: -1.6074

Max Act: 1.5680. Min Act: -0.0001

Data Index: 406250 (C4 (Web Text))

Max Activating Token Index: 395

Text #2

Max Range: 1.6074. Min Range: -1.6074

Max Act: 1.2294. Min Act: -0.0001

Data Index: 231309 (C4 (Web Text))

Max Activating Token Index: 139

Text #3

Max Range: 1.6074. Min Range: -1.6074

Max Act: 1.3227. Min Act: -0.0001

Data Index: 419209 (C4 (Web Text))

Max Activating Token Index: 454

Text #4

Max Range: 1.6074. Min Range: -1.6074

Max Act: 1.3932. Min Act: -0.0001

Data Index: 241270 (C4 (Web Text))

Max Activating Token Index: 320

Text #5

Max Range: 1.6074. Min Range: -1.6074

Max Act: 1.1471. Min Act: -0.0001

Data Index: 282259 (C4 (Web Text))

Max Activating Token Index: 169

Text #6

Max Range: 1.6074. Min Range: -1.6074

Max Act: 1.2849. Min Act: -0.0001

Data Index: 434820 (C4 (Web Text))

Max Activating Token Index: 622

Text #7

Max Range: 1.6074. Min Range: -1.6074

Max Act: 1.1255. Min Act: -0.0001

Data Index: 961690 (C4 (Web Text))

Max Activating Token Index: 5

Text #8

Max Range: 1.6074. Min Range: -1.6074

Max Act: 1.1481. Min Act: -0.0001

Data Index: 755770 (C4 (Web Text))

Max Activating Token Index: 866

Text #9

Max Range: 1.6074. Min Range: -1.6074

Max Act: 1.1473. Min Act: -0.0001

Data Index: 263395 (C4 (Web Text))

Max Activating Token Index: 118

Text #10

Max Range: 1.6074. Min Range: -1.6074

Max Act: 0.8683. Min Act: -0.0001

Data Index: 981752 (C4 (Web Text))

Max Activating Token Index: 700

Text #11

Max Range: 1.6074. Min Range: -1.6074

Max Act: 0.7958. Min Act: -0.0001

Data Index: 959514 (C4 (Web Text))

Max Activating Token Index: 914

Text #12

Max Range: 1.6074. Min Range: -1.6074

Max Act: 0.8496. Min Act: -0.0001

Data Index: 456704 (C4 (Web Text))

Max Activating Token Index: 608

Text #13

Max Range: 1.6074. Min Range: -1.6074

Max Act: 0.9164. Min Act: -0.0001

Data Index: 1545893 (Python Code)

Max Activating Token Index: 626

Text #14

Max Range: 1.6074. Min Range: -1.6074

Max Act: 0.8740. Min Act: -0.0001

Data Index: 421907 (C4 (Web Text))

Max Activating Token Index: 834

Text #15

Max Range: 1.6074. Min Range: -1.6074

Max Act: 0.8506. Min Act: -0.0001

Data Index: 4003 (C4 (Web Text))

Max Activating Token Index: 439

Text #16

Max Range: 1.6074. Min Range: -1.6074

Max Act: 0.7907. Min Act: -0.0001

Data Index: 1558913 (Python Code)

Max Activating Token Index: 408

Text #17

Max Range: 1.6074. Min Range: -1.6074

Max Act: 0.7102. Min Act: -0.0001

Data Index: 522255 (C4 (Web Text))

Max Activating Token Index: 273

Text #18

Max Range: 1.6074. Min Range: -1.6074

Max Act: 0.7715. Min Act: -0.0001

Data Index: 1247629 (C4 (Web Text))

Max Activating Token Index: 583

Text #19

Max Range: 1.6074. Min Range: -1.6074

Max Act: 0.7974. Min Act: -0.0001

Data Index: 869054 (C4 (Web Text))

Max Activating Token Index: 942