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

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

Neuron 5533 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.3933. Min Range: -2.3933

Max Act: 2.3933. Min Act: -0.0000

Data Index: 941748 (C4 (Web Text))

Max Activating Token Index: 584

Text #1

Max Range: 2.3933. Min Range: -2.3933

Max Act: 2.3370. Min Act: -0.0000

Data Index: 1630766 (Python Code)

Max Activating Token Index: 799

Text #2

Max Range: 2.3933. Min Range: -2.3933

Max Act: 2.1531. Min Act: -0.0000

Data Index: 653430 (C4 (Web Text))

Max Activating Token Index: 863

Text #3

Max Range: 2.3933. Min Range: -2.3933

Max Act: 2.0064. Min Act: -0.0000

Data Index: 1021828 (C4 (Web Text))

Max Activating Token Index: 330

Text #4

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.8537. Min Act: -0.0000

Data Index: 831856 (C4 (Web Text))

Max Activating Token Index: 840

Text #5

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.8380. Min Act: -0.0000

Data Index: 1324724 (C4 (Web Text))

Max Activating Token Index: 324

Text #6

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.7352. Min Act: -0.0000

Data Index: 698851 (C4 (Web Text))

Max Activating Token Index: 721

Text #7

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.8246. Min Act: -0.0000

Data Index: 99889 (C4 (Web Text))

Max Activating Token Index: 247

Text #8

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.8140. Min Act: -0.0000

Data Index: 1304638 (C4 (Web Text))

Max Activating Token Index: 855

Text #9

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.9186. Min Act: -0.0000

Data Index: 128654 (C4 (Web Text))

Max Activating Token Index: 312

Text #10

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.8841. Min Act: -0.0000

Data Index: 1552625 (Python Code)

Max Activating Token Index: 485

Text #11

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.8753. Min Act: -0.0000

Data Index: 655122 (C4 (Web Text))

Max Activating Token Index: 84

Text #12

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.6766. Min Act: -0.0000

Data Index: 1172974 (C4 (Web Text))

Max Activating Token Index: 381

Text #13

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.6659. Min Act: -0.0000

Data Index: 822604 (C4 (Web Text))

Max Activating Token Index: 388

Text #14

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.8347. Min Act: -0.0000

Data Index: 277955 (C4 (Web Text))

Max Activating Token Index: 671

Text #15

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.8011. Min Act: -0.0000

Data Index: 87794 (C4 (Web Text))

Max Activating Token Index: 796

Text #16

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.7964. Min Act: -0.0000

Data Index: 728168 (C4 (Web Text))

Max Activating Token Index: 868

Text #17

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.7825. Min Act: -0.0000

Data Index: 1404235 (Python Code)

Max Activating Token Index: 699

Text #18

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.7644. Min Act: -0.0000

Data Index: 1346726 (C4 (Web Text))

Max Activating Token Index: 494

Text #19

Max Range: 2.3933. Min Range: -2.3933

Max Act: 1.7635. Min Act: -0.0000

Data Index: 662642 (C4 (Web Text))

Max Activating Token Index: 601