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

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

Neuron 3928 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: 0.6996. Min Range: -0.6996

Max Act: 0.6787. Min Act: -0.0000

Data Index: 1374112 (Python Code)

Max Activating Token Index: 754

Text #1

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.6774. Min Act: -0.0000

Data Index: 547944 (C4 (Web Text))

Max Activating Token Index: 429

Text #2

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.6996. Min Act: -0.0000

Data Index: 1004705 (C4 (Web Text))

Max Activating Token Index: 107

Text #3

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.6166. Min Act: -0.0000

Data Index: 162172 (C4 (Web Text))

Max Activating Token Index: 636

Text #4

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.6194. Min Act: -0.0000

Data Index: 1632154 (Python Code)

Max Activating Token Index: 550

Text #5

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.6328. Min Act: -0.0000

Data Index: 405257 (C4 (Web Text))

Max Activating Token Index: 717

Text #6

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.5294. Min Act: -0.0000

Data Index: 170980 (C4 (Web Text))

Max Activating Token Index: 246

Text #7

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.5305. Min Act: -0.0000

Data Index: 1380604 (Python Code)

Max Activating Token Index: 518

Text #8

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.5077. Min Act: -0.0000

Data Index: 80900 (C4 (Web Text))

Max Activating Token Index: 876

Text #9

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.4912. Min Act: -0.0000

Data Index: 1615200 (Python Code)

Max Activating Token Index: 944

Text #10

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.4966. Min Act: -0.0000

Data Index: 580431 (C4 (Web Text))

Max Activating Token Index: 714

Text #11

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.4919. Min Act: -0.0000

Data Index: 24009 (C4 (Web Text))

Max Activating Token Index: 10

Text #12

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.4827. Min Act: -0.0000

Data Index: 1626304 (Python Code)

Max Activating Token Index: 111

Text #13

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.4642. Min Act: -0.0000

Data Index: 355736 (C4 (Web Text))

Max Activating Token Index: 909

Text #14

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.4863. Min Act: -0.0000

Data Index: 125598 (C4 (Web Text))

Max Activating Token Index: 730

Text #15

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.4880. Min Act: -0.0000

Data Index: 906599 (C4 (Web Text))

Max Activating Token Index: 335

Text #16

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.4611. Min Act: -0.0000

Data Index: 702877 (C4 (Web Text))

Max Activating Token Index: 107

Text #17

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.5709. Min Act: -0.0000

Data Index: 1024301 (C4 (Web Text))

Max Activating Token Index: 233

Text #18

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.4953. Min Act: -0.0000

Data Index: 1465308 (Python Code)

Max Activating Token Index: 313

Text #19

Max Range: 0.6996. Min Range: -0.6996

Max Act: 0.4538. Min Act: -0.0000

Data Index: 1010628 (C4 (Web Text))

Max Activating Token Index: 833