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

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

Neuron 4839 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.2404. Min Range: -1.2404

Max Act: 1.0884. Min Act: -0.0000

Data Index: 1154796 (C4 (Web Text))

Max Activating Token Index: 663

Text #1

Max Range: 1.2404. Min Range: -1.2404

Max Act: 1.2404. Min Act: -0.0000

Data Index: 587155 (C4 (Web Text))

Max Activating Token Index: 385

Text #2

Max Range: 1.2404. Min Range: -1.2404

Max Act: 1.0602. Min Act: -0.0000

Data Index: 234604 (C4 (Web Text))

Max Activating Token Index: 885

Text #3

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.9373. Min Act: -0.0000

Data Index: 424556 (C4 (Web Text))

Max Activating Token Index: 984

Text #4

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.8639. Min Act: -0.0000

Data Index: 794023 (C4 (Web Text))

Max Activating Token Index: 648

Text #5

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.8104. Min Act: -0.0000

Data Index: 615240 (C4 (Web Text))

Max Activating Token Index: 496

Text #6

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.8448. Min Act: -0.0000

Data Index: 1093259 (C4 (Web Text))

Max Activating Token Index: 380

Text #7

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.7245. Min Act: -0.0000

Data Index: 1089843 (C4 (Web Text))

Max Activating Token Index: 777

Text #8

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.6974. Min Act: -0.0000

Data Index: 208436 (C4 (Web Text))

Max Activating Token Index: 506

Text #9

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.6306. Min Act: 0.0000

Data Index: 1363805 (Python Code)

Max Activating Token Index: 348

Text #10

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.6354. Min Act: -0.0000

Data Index: 635955 (C4 (Web Text))

Max Activating Token Index: 593

Text #11

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.6793. Min Act: -0.0000

Data Index: 582197 (C4 (Web Text))

Max Activating Token Index: 426

Text #12

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.6007. Min Act: -0.0000

Data Index: 3526 (C4 (Web Text))

Max Activating Token Index: 718

Text #13

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.5710. Min Act: -0.0000

Data Index: 1549593 (Python Code)

Max Activating Token Index: 746

Text #14

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.5577. Min Act: -0.0000

Data Index: 1277048 (C4 (Web Text))

Max Activating Token Index: 876

Text #15

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.5523. Min Act: -0.0000

Data Index: 1498535 (Python Code)

Max Activating Token Index: 1006

Text #16

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.5517. Min Act: -0.0000

Data Index: 1357482 (C4 (Web Text))

Max Activating Token Index: 852

Text #17

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.5700. Min Act: -0.0000

Data Index: 1430665 (Python Code)

Max Activating Token Index: 123

Text #18

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.5285. Min Act: -0.0000

Data Index: 1591210 (Python Code)

Max Activating Token Index: 845

Text #19

Max Range: 1.2404. Min Range: -1.2404

Max Act: 0.5331. Min Act: -0.0000

Data Index: 953597 (C4 (Web Text))

Max Activating Token Index: 727