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

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

Neuron 3750 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.7396. Min Act: -0.0000

Data Index: 1557564 (Python Code)

Max Activating Token Index: 871

Text #1

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.5967. Min Act: -0.0000

Data Index: 1500883 (Python Code)

Max Activating Token Index: 585

Text #2

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.7027. Min Act: -0.0000

Data Index: 1565227 (Python Code)

Max Activating Token Index: 520

Text #3

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.6901. Min Act: -0.0000

Data Index: 15548 (C4 (Web Text))

Max Activating Token Index: 94

Text #4

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.8287. Min Act: -0.0000

Data Index: 1263470 (C4 (Web Text))

Max Activating Token Index: 771

Text #5

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.5540. Min Act: -0.0000

Data Index: 1522301 (Python Code)

Max Activating Token Index: 692

Text #6

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.6869. Min Act: -0.0000

Data Index: 216276 (C4 (Web Text))

Max Activating Token Index: 356

Text #7

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.5613. Min Act: -0.0000

Data Index: 154939 (C4 (Web Text))

Max Activating Token Index: 370

Text #8

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.5300. Min Act: -0.0000

Data Index: 127843 (C4 (Web Text))

Max Activating Token Index: 165

Text #9

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.1239. Min Act: -0.0000

Data Index: 42814 (C4 (Web Text))

Max Activating Token Index: 209

Text #10

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.4726. Min Act: -0.0000

Data Index: 373808 (C4 (Web Text))

Max Activating Token Index: 123

Text #11

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.8376. Min Act: -0.0000

Data Index: 656880 (C4 (Web Text))

Max Activating Token Index: 821

Text #12

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.5736. Min Act: -0.0000

Data Index: 1639601 (Python Code)

Max Activating Token Index: 582

Text #13

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.5651. Min Act: -0.0000

Data Index: 5136 (C4 (Web Text))

Max Activating Token Index: 936

Text #14

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.3857. Min Act: -0.0000

Data Index: 1258393 (C4 (Web Text))

Max Activating Token Index: 674

Text #15

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.5139. Min Act: -0.0000

Data Index: 960990 (C4 (Web Text))

Max Activating Token Index: 449

Text #16

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.4779. Min Act: -0.0000

Data Index: 997472 (C4 (Web Text))

Max Activating Token Index: 429

Text #17

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.3542. Min Act: -0.0000

Data Index: 650696 (C4 (Web Text))

Max Activating Token Index: 865

Text #18

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.3348. Min Act: -0.0000

Data Index: 1218965 (C4 (Web Text))

Max Activating Token Index: 158

Text #19

Max Range: 2.8376. Min Range: -2.8376

Max Act: 2.5813. Min Act: -0.0000

Data Index: 1272174 (C4 (Web Text))

Max Activating Token Index: 354