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

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

Neuron 560 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.1722. Min Range: -1.1722

Max Act: 1.1722. Min Act: -0.0000

Data Index: 1201644 (C4 (Web Text))

Max Activating Token Index: 510

Text #1

Max Range: 1.1722. Min Range: -1.1722

Max Act: 1.1174. Min Act: -0.0000

Data Index: 312299 (C4 (Web Text))

Max Activating Token Index: 566

Text #2

Max Range: 1.1722. Min Range: -1.1722

Max Act: 1.1480. Min Act: -0.0000

Data Index: 95300 (C4 (Web Text))

Max Activating Token Index: 718

Text #3

Max Range: 1.1722. Min Range: -1.1722

Max Act: 1.1247. Min Act: -0.0000

Data Index: 470689 (C4 (Web Text))

Max Activating Token Index: 458

Text #4

Max Range: 1.1722. Min Range: -1.1722

Max Act: 1.0287. Min Act: -0.0000

Data Index: 1551269 (Python Code)

Max Activating Token Index: 416

Text #5

Max Range: 1.1722. Min Range: -1.1722

Max Act: 1.0264. Min Act: -0.0000

Data Index: 543553 (C4 (Web Text))

Max Activating Token Index: 195

Text #6

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.9859. Min Act: -0.0000

Data Index: 460282 (C4 (Web Text))

Max Activating Token Index: 924

Text #7

Max Range: 1.1722. Min Range: -1.1722

Max Act: 1.0323. Min Act: -0.0000

Data Index: 1045577 (C4 (Web Text))

Max Activating Token Index: 400

Text #8

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.9615. Min Act: -0.0000

Data Index: 401653 (C4 (Web Text))

Max Activating Token Index: 15

Text #9

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.9368. Min Act: -0.0000

Data Index: 822413 (C4 (Web Text))

Max Activating Token Index: 637

Text #10

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.9646. Min Act: -0.0000

Data Index: 1228738 (C4 (Web Text))

Max Activating Token Index: 653

Text #11

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.9885. Min Act: -0.0000

Data Index: 1167860 (C4 (Web Text))

Max Activating Token Index: 292

Text #12

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.9520. Min Act: -0.0000

Data Index: 88490 (C4 (Web Text))

Max Activating Token Index: 10

Text #13

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.9115. Min Act: -0.0000

Data Index: 1009811 (C4 (Web Text))

Max Activating Token Index: 357

Text #14

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.9023. Min Act: -0.0000

Data Index: 1529372 (Python Code)

Max Activating Token Index: 511

Text #15

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.8515. Min Act: -0.0000

Data Index: 61126 (C4 (Web Text))

Max Activating Token Index: 762

Text #16

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.8707. Min Act: -0.0000

Data Index: 771857 (C4 (Web Text))

Max Activating Token Index: 83

Text #17

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.8828. Min Act: -0.0000

Data Index: 606754 (C4 (Web Text))

Max Activating Token Index: 232

Text #18

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.8519. Min Act: -0.0000

Data Index: 747189 (C4 (Web Text))

Max Activating Token Index: 756

Text #19

Max Range: 1.1722. Min Range: -1.1722

Max Act: 0.8683. Min Act: -0.0000

Data Index: 885047 (C4 (Web Text))

Max Activating Token Index: 21