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

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

Neuron 3358 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: 4.1948. Min Range: -4.1948

Max Act: 3.9108. Min Act: -0.0000

Data Index: 800396 (C4 (Web Text))

Max Activating Token Index: 39

Text #1

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.7650. Min Act: -0.0000

Data Index: 1514531 (Python Code)

Max Activating Token Index: 146

Text #2

Max Range: 4.1948. Min Range: -4.1948

Max Act: 4.1948. Min Act: -0.0000

Data Index: 1452142 (Python Code)

Max Activating Token Index: 478

Text #3

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.6541. Min Act: -0.0000

Data Index: 501778 (C4 (Web Text))

Max Activating Token Index: 49

Text #4

Max Range: 4.1948. Min Range: -4.1948

Max Act: 2.7443. Min Act: -0.0000

Data Index: 1372924 (Python Code)

Max Activating Token Index: 261

Text #5

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.0282. Min Act: -0.0000

Data Index: 955920 (C4 (Web Text))

Max Activating Token Index: 923

Text #6

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.3372. Min Act: -0.0000

Data Index: 40007 (C4 (Web Text))

Max Activating Token Index: 57

Text #7

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.3463. Min Act: -0.0000

Data Index: 1389604 (Python Code)

Max Activating Token Index: 782

Text #8

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.0320. Min Act: -0.0000

Data Index: 620349 (C4 (Web Text))

Max Activating Token Index: 964

Text #9

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.2110. Min Act: -0.0000

Data Index: 889949 (C4 (Web Text))

Max Activating Token Index: 1023

Text #10

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.1731. Min Act: -0.0000

Data Index: 703637 (C4 (Web Text))

Max Activating Token Index: 105

Text #11

Max Range: 4.1948. Min Range: -4.1948

Max Act: 2.8908. Min Act: -0.0000

Data Index: 1469394 (Python Code)

Max Activating Token Index: 813

Text #12

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.1928. Min Act: -0.0000

Data Index: 962409 (C4 (Web Text))

Max Activating Token Index: 849

Text #13

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.1339. Min Act: -0.0000

Data Index: 710904 (C4 (Web Text))

Max Activating Token Index: 843

Text #14

Max Range: 4.1948. Min Range: -4.1948

Max Act: 2.9798. Min Act: -0.0000

Data Index: 565289 (C4 (Web Text))

Max Activating Token Index: 135

Text #15

Max Range: 4.1948. Min Range: -4.1948

Max Act: 2.9707. Min Act: -0.0000

Data Index: 539388 (C4 (Web Text))

Max Activating Token Index: 141

Text #16

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.2659. Min Act: -0.0000

Data Index: 1654841 (Python Code)

Max Activating Token Index: 685

Text #17

Max Range: 4.1948. Min Range: -4.1948

Max Act: 2.7992. Min Act: -0.0000

Data Index: 891121 (C4 (Web Text))

Max Activating Token Index: 924

Text #18

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.1082. Min Act: -0.0000

Data Index: 687813 (C4 (Web Text))

Max Activating Token Index: 50

Text #19

Max Range: 4.1948. Min Range: -4.1948

Max Act: 3.2477. Min Act: -0.0000

Data Index: 306626 (C4 (Web Text))

Max Activating Token Index: 849