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

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

Neuron 3724 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.2465. Min Act: -0.0001

Data Index: 1113537 (C4 (Web Text))

Max Activating Token Index: 466

Text #1

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1962. Min Act: -0.0000

Data Index: 896330 (C4 (Web Text))

Max Activating Token Index: 705

Text #2

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1497. Min Act: -0.0000

Data Index: 156534 (C4 (Web Text))

Max Activating Token Index: 257

Text #3

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.2111. Min Act: -0.0000

Data Index: 1354579 (C4 (Web Text))

Max Activating Token Index: 456

Text #4

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1831. Min Act: -0.0000

Data Index: 229572 (C4 (Web Text))

Max Activating Token Index: 799

Text #5

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.2103. Min Act: -0.0000

Data Index: 532581 (C4 (Web Text))

Max Activating Token Index: 879

Text #6

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1903. Min Act: -0.0000

Data Index: 1241094 (C4 (Web Text))

Max Activating Token Index: 70

Text #7

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1883. Min Act: -0.0000

Data Index: 1117512 (C4 (Web Text))

Max Activating Token Index: 541

Text #8

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.2068. Min Act: -0.0000

Data Index: 1558257 (Python Code)

Max Activating Token Index: 885

Text #9

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.2081. Min Act: -0.0000

Data Index: 1058921 (C4 (Web Text))

Max Activating Token Index: 184

Text #10

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1639. Min Act: -0.0000

Data Index: 1198975 (C4 (Web Text))

Max Activating Token Index: 203

Text #11

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1801. Min Act: -0.0000

Data Index: 991829 (C4 (Web Text))

Max Activating Token Index: 967

Text #12

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1925. Min Act: -0.0000

Data Index: 864305 (C4 (Web Text))

Max Activating Token Index: 241

Text #13

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1928. Min Act: -0.0000

Data Index: 70743 (C4 (Web Text))

Max Activating Token Index: 932

Text #14

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1665. Min Act: -0.0000

Data Index: 559078 (C4 (Web Text))

Max Activating Token Index: 749

Text #15

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1782. Min Act: -0.0000

Data Index: 1639998 (Python Code)

Max Activating Token Index: 655

Text #16

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1805. Min Act: -0.0001

Data Index: 213429 (C4 (Web Text))

Max Activating Token Index: 207

Text #17

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1771. Min Act: -0.0001

Data Index: 710644 (C4 (Web Text))

Max Activating Token Index: 421

Text #18

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1821. Min Act: -0.0001

Data Index: 108834 (C4 (Web Text))

Max Activating Token Index: 564

Text #19

Max Range: 0.2465. Min Range: -0.2465

Max Act: 0.1814. Min Act: -0.0000

Data Index: 119564 (C4 (Web Text))

Max Activating Token Index: 548