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

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

Neuron 1759 in Layer 11

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.1040. Min Act: -0.0000

Data Index: 467199 (C4 (Web Text))

Max Activating Token Index: 265

Text #1

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0983. Min Act: -0.0000

Data Index: 1306142 (C4 (Web Text))

Max Activating Token Index: 688

Text #2

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0794. Min Act: -0.0000

Data Index: 40763 (C4 (Web Text))

Max Activating Token Index: 97

Text #3

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0818. Min Act: -0.0000

Data Index: 326365 (C4 (Web Text))

Max Activating Token Index: 393

Text #4

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0785. Min Act: -0.0000

Data Index: 253242 (C4 (Web Text))

Max Activating Token Index: 571

Text #5

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0805. Min Act: 0.0000

Data Index: 1385780 (Python Code)

Max Activating Token Index: 326

Text #6

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0770. Min Act: -0.0000

Data Index: 756677 (C4 (Web Text))

Max Activating Token Index: 20

Text #7

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0831. Min Act: -0.0000

Data Index: 1352025 (C4 (Web Text))

Max Activating Token Index: 724

Text #8

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0738. Min Act: -0.0000

Data Index: 196994 (C4 (Web Text))

Max Activating Token Index: 66

Text #9

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0761. Min Act: -0.0000

Data Index: 1375876 (Python Code)

Max Activating Token Index: 803

Text #10

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0686. Min Act: -0.0000

Data Index: 910124 (C4 (Web Text))

Max Activating Token Index: 1021

Text #11

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0654. Min Act: -0.0000

Data Index: 1393619 (Python Code)

Max Activating Token Index: 659

Text #12

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0673. Min Act: -0.0000

Data Index: 1079664 (C4 (Web Text))

Max Activating Token Index: 827

Text #13

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0679. Min Act: -0.0000

Data Index: 529917 (C4 (Web Text))

Max Activating Token Index: 262

Text #14

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0640. Min Act: -0.0000

Data Index: 1483757 (Python Code)

Max Activating Token Index: 72

Text #15

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0646. Min Act: 0.0000

Data Index: 1430614 (Python Code)

Max Activating Token Index: 478

Text #16

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0633. Min Act: -0.0000

Data Index: 872575 (C4 (Web Text))

Max Activating Token Index: 94

Text #17

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0636. Min Act: -0.0000

Data Index: 1498600 (Python Code)

Max Activating Token Index: 431

Text #18

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0585. Min Act: -0.0000

Data Index: 66557 (C4 (Web Text))

Max Activating Token Index: 294

Text #19

Max Range: 0.1040. Min Range: -0.1040

Max Act: 0.0642. Min Act: -0.0000

Data Index: 767672 (C4 (Web Text))

Max Activating Token Index: 1020