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

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

Neuron 3041 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.8633. Min Act: -0.0000

Data Index: 17920 (C4 (Web Text))

Max Activating Token Index: 647

Text #1

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.7527. Min Act: -0.0000

Data Index: 1572964 (Python Code)

Max Activating Token Index: 129

Text #2

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.5099. Min Act: -0.0000

Data Index: 1275792 (C4 (Web Text))

Max Activating Token Index: 66

Text #3

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.5353. Min Act: -0.0000

Data Index: 173956 (C4 (Web Text))

Max Activating Token Index: 607

Text #4

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.4448. Min Act: -0.0000

Data Index: 737643 (C4 (Web Text))

Max Activating Token Index: 113

Text #5

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.5233. Min Act: -0.0000

Data Index: 441766 (C4 (Web Text))

Max Activating Token Index: 797

Text #6

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.4643. Min Act: -0.0000

Data Index: 285683 (C4 (Web Text))

Max Activating Token Index: 13

Text #7

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.4144. Min Act: -0.0000

Data Index: 331491 (C4 (Web Text))

Max Activating Token Index: 855

Text #8

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.4102. Min Act: -0.0000

Data Index: 1282263 (C4 (Web Text))

Max Activating Token Index: 655

Text #9

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.3443. Min Act: -0.0000

Data Index: 907825 (C4 (Web Text))

Max Activating Token Index: 627

Text #10

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.3218. Min Act: -0.0000

Data Index: 1375090 (Python Code)

Max Activating Token Index: 613

Text #11

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.3381. Min Act: -0.0000

Data Index: 1525467 (Python Code)

Max Activating Token Index: 935

Text #12

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.2886. Min Act: -0.0000

Data Index: 955357 (C4 (Web Text))

Max Activating Token Index: 350

Text #13

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.3176. Min Act: -0.0000

Data Index: 644120 (C4 (Web Text))

Max Activating Token Index: 136

Text #14

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.3578. Min Act: -0.0000

Data Index: 1489566 (Python Code)

Max Activating Token Index: 527

Text #15

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.2865. Min Act: -0.0000

Data Index: 1280208 (C4 (Web Text))

Max Activating Token Index: 836

Text #16

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.2408. Min Act: -0.0000

Data Index: 766152 (C4 (Web Text))

Max Activating Token Index: 343

Text #17

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.1853. Min Act: -0.0000

Data Index: 913368 (C4 (Web Text))

Max Activating Token Index: 898

Text #18

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.2481. Min Act: -0.0000

Data Index: 1479320 (Python Code)

Max Activating Token Index: 427

Text #19

Max Range: 1.8633. Min Range: -1.8633

Max Act: 1.2144. Min Act: -0.0000

Data Index: 356497 (C4 (Web Text))

Max Activating Token Index: 278