Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 237 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0838. Min Act: -0.0001

Data Index: 897980 (C4 (Web Text))

Max Activating Token Index: 664

Text #1

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0673. Min Act: -0.0002

Data Index: 54804 (C4 (Web Text))

Max Activating Token Index: 89

Text #2

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0634. Min Act: -0.0002

Data Index: 1343323 (C4 (Web Text))

Max Activating Token Index: 555

Text #3

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0629. Min Act: -0.0002

Data Index: 47930 (C4 (Web Text))

Max Activating Token Index: 421

Text #4

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0634. Min Act: -0.0002

Data Index: 674957 (C4 (Web Text))

Max Activating Token Index: 864

Text #5

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0603. Min Act: -0.0002

Data Index: 165011 (C4 (Web Text))

Max Activating Token Index: 397

Text #6

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0545. Min Act: -0.0002

Data Index: 1389560 (Python Code)

Max Activating Token Index: 412

Text #7

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0555. Min Act: -0.0002

Data Index: 1246511 (C4 (Web Text))

Max Activating Token Index: 457

Text #8

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0554. Min Act: -0.0002

Data Index: 947271 (C4 (Web Text))

Max Activating Token Index: 918

Text #9

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0529. Min Act: -0.0002

Data Index: 407836 (C4 (Web Text))

Max Activating Token Index: 796

Text #10

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0540. Min Act: -0.0002

Data Index: 1643254 (Python Code)

Max Activating Token Index: 877

Text #11

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0522. Min Act: -0.0002

Data Index: 1347354 (C4 (Web Text))

Max Activating Token Index: 891

Text #12

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0500. Min Act: -0.0002

Data Index: 675150 (C4 (Web Text))

Max Activating Token Index: 924

Text #13

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0500. Min Act: -0.0001

Data Index: 1169186 (C4 (Web Text))

Max Activating Token Index: 517

Text #14

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0531. Min Act: -0.0001

Data Index: 1553257 (Python Code)

Max Activating Token Index: 632

Text #15

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0505. Min Act: -0.0001

Data Index: 1626164 (Python Code)

Max Activating Token Index: 853

Text #16

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0498. Min Act: -0.0002

Data Index: 1456514 (Python Code)

Max Activating Token Index: 679

Text #17

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0499. Min Act: -0.0002

Data Index: 658737 (C4 (Web Text))

Max Activating Token Index: 266

Text #18

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0479. Min Act: -0.0002

Data Index: 38141 (C4 (Web Text))

Max Activating Token Index: 536

Text #19

Max Range: 0.0838. Min Range: -0.0838

Max Act: 0.0497. Min Act: -0.0001

Data Index: 1593853 (Python Code)

Max Activating Token Index: 829