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

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

Neuron 424 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: 0.1685. Min Range: -0.1685

Max Act: 0.1641. Min Act: -0.0002

Data Index: 936506 (C4 (Web Text))

Max Activating Token Index: 727

Text #1

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1685. Min Act: -0.0002

Data Index: 1348564 (C4 (Web Text))

Max Activating Token Index: 26

Text #2

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1673. Min Act: -0.0002

Data Index: 1226427 (C4 (Web Text))

Max Activating Token Index: 11

Text #3

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1545. Min Act: -0.0001

Data Index: 519701 (C4 (Web Text))

Max Activating Token Index: 10

Text #4

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1597. Min Act: -0.0002

Data Index: 81414 (C4 (Web Text))

Max Activating Token Index: 54

Text #5

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1634. Min Act: -0.0002

Data Index: 1287289 (C4 (Web Text))

Max Activating Token Index: 32

Text #6

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1560. Min Act: -0.0002

Data Index: 1100568 (C4 (Web Text))

Max Activating Token Index: 33

Text #7

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1581. Min Act: -0.0002

Data Index: 417458 (C4 (Web Text))

Max Activating Token Index: 996

Text #8

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1561. Min Act: -0.0002

Data Index: 375840 (C4 (Web Text))

Max Activating Token Index: 123

Text #9

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1518. Min Act: -0.0002

Data Index: 124191 (C4 (Web Text))

Max Activating Token Index: 31

Text #10

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1531. Min Act: -0.0002

Data Index: 352339 (C4 (Web Text))

Max Activating Token Index: 475

Text #11

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1525. Min Act: -0.0002

Data Index: 843533 (C4 (Web Text))

Max Activating Token Index: 25

Text #12

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1515. Min Act: -0.0002

Data Index: 39472 (C4 (Web Text))

Max Activating Token Index: 392

Text #13

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1477. Min Act: -0.0002

Data Index: 792202 (C4 (Web Text))

Max Activating Token Index: 770

Text #14

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1503. Min Act: -0.0002

Data Index: 288085 (C4 (Web Text))

Max Activating Token Index: 17

Text #15

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1438. Min Act: -0.0002

Data Index: 1380357 (Python Code)

Max Activating Token Index: 471

Text #16

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1404. Min Act: -0.0002

Data Index: 513594 (C4 (Web Text))

Max Activating Token Index: 753

Text #17

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1465. Min Act: -0.0002

Data Index: 966314 (C4 (Web Text))

Max Activating Token Index: 134

Text #18

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1409. Min Act: -0.0002

Data Index: 1537493 (Python Code)

Max Activating Token Index: 638

Text #19

Max Range: 0.1685. Min Range: -0.1685

Max Act: 0.1498. Min Act: -0.0002

Data Index: 989346 (C4 (Web Text))

Max Activating Token Index: 966