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

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

Neuron 3913 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.9309. Min Act: -0.0000

Data Index: 941591 (C4 (Web Text))

Max Activating Token Index: 347

Text #1

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.8313. Min Act: -0.0000

Data Index: 1132685 (C4 (Web Text))

Max Activating Token Index: 102

Text #2

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7839. Min Act: -0.0000

Data Index: 630797 (C4 (Web Text))

Max Activating Token Index: 250

Text #3

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7785. Min Act: -0.0000

Data Index: 1474646 (Python Code)

Max Activating Token Index: 823

Text #4

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7978. Min Act: -0.0000

Data Index: 589492 (C4 (Web Text))

Max Activating Token Index: 89

Text #5

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.6807. Min Act: -0.0000

Data Index: 693662 (C4 (Web Text))

Max Activating Token Index: 335

Text #6

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7707. Min Act: -0.0000

Data Index: 1508240 (Python Code)

Max Activating Token Index: 581

Text #7

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.6834. Min Act: -0.0000

Data Index: 855336 (C4 (Web Text))

Max Activating Token Index: 581

Text #8

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7693. Min Act: -0.0000

Data Index: 52849 (C4 (Web Text))

Max Activating Token Index: 410

Text #9

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7587. Min Act: -0.0000

Data Index: 1408423 (Python Code)

Max Activating Token Index: 905

Text #10

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7999. Min Act: -0.0001

Data Index: 1005715 (C4 (Web Text))

Max Activating Token Index: 271

Text #11

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7680. Min Act: -0.0000

Data Index: 1654874 (Python Code)

Max Activating Token Index: 766

Text #12

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7684. Min Act: -0.0000

Data Index: 1473385 (Python Code)

Max Activating Token Index: 873

Text #13

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7459. Min Act: -0.0001

Data Index: 1556649 (Python Code)

Max Activating Token Index: 495

Text #14

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7480. Min Act: -0.0000

Data Index: 1471292 (Python Code)

Max Activating Token Index: 818

Text #15

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7433. Min Act: -0.0001

Data Index: 1602138 (Python Code)

Max Activating Token Index: 130

Text #16

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7832. Min Act: -0.0000

Data Index: 1422676 (Python Code)

Max Activating Token Index: 100

Text #17

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7561. Min Act: -0.0000

Data Index: 1433377 (Python Code)

Max Activating Token Index: 74

Text #18

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7798. Min Act: -0.0000

Data Index: 472472 (C4 (Web Text))

Max Activating Token Index: 212

Text #19

Max Range: 0.9309. Min Range: -0.9309

Max Act: 0.7350. Min Act: -0.0000

Data Index: 1560622 (Python Code)

Max Activating Token Index: 332