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

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

Neuron 3549 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0710. Min Act: -0.0001

Data Index: 477741 (C4 (Web Text))

Max Activating Token Index: 661

Text #1

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0704. Min Act: -0.0001

Data Index: 1220530 (C4 (Web Text))

Max Activating Token Index: 980

Text #2

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0619. Min Act: -0.0001

Data Index: 102150 (C4 (Web Text))

Max Activating Token Index: 803

Text #3

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0667. Min Act: -0.0001

Data Index: 1353431 (C4 (Web Text))

Max Activating Token Index: 375

Text #4

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0618. Min Act: -0.0001

Data Index: 190384 (C4 (Web Text))

Max Activating Token Index: 285

Text #5

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0610. Min Act: -0.0001

Data Index: 1622457 (Python Code)

Max Activating Token Index: 491

Text #6

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0626. Min Act: -0.0001

Data Index: 1610892 (Python Code)

Max Activating Token Index: 89

Text #7

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0574. Min Act: -0.0001

Data Index: 158936 (C4 (Web Text))

Max Activating Token Index: 1007

Text #8

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0641. Min Act: -0.0001

Data Index: 54825 (C4 (Web Text))

Max Activating Token Index: 218

Text #9

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0610. Min Act: -0.0001

Data Index: 1651773 (Python Code)

Max Activating Token Index: 843

Text #10

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0588. Min Act: -0.0001

Data Index: 10514 (C4 (Web Text))

Max Activating Token Index: 494

Text #11

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0561. Min Act: -0.0001

Data Index: 1638459 (Python Code)

Max Activating Token Index: 195

Text #12

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0559. Min Act: -0.0001

Data Index: 996620 (C4 (Web Text))

Max Activating Token Index: 60

Text #13

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0570. Min Act: -0.0001

Data Index: 1465790 (Python Code)

Max Activating Token Index: 789

Text #14

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0580. Min Act: -0.0001

Data Index: 1570620 (Python Code)

Max Activating Token Index: 584

Text #15

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0579. Min Act: -0.0001

Data Index: 1031684 (C4 (Web Text))

Max Activating Token Index: 114

Text #16

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0604. Min Act: -0.0001

Data Index: 1455685 (Python Code)

Max Activating Token Index: 175

Text #17

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0550. Min Act: -0.0001

Data Index: 817559 (C4 (Web Text))

Max Activating Token Index: 739

Text #18

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0570. Min Act: -0.0001

Data Index: 47983 (C4 (Web Text))

Max Activating Token Index: 818

Text #19

Max Range: 0.0710. Min Range: -0.0710

Max Act: 0.0512. Min Act: -0.0001

Data Index: 284858 (C4 (Web Text))

Max Activating Token Index: 685