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

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

Neuron 1831 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.4508. Min Range: -3.4508

Max Act: 3.4508. Min Act: -0.0002

Data Index: 1416177 (Python Code)

Max Activating Token Index: 5

Text #1

Max Range: 3.4508. Min Range: -3.4508

Max Act: 3.3286. Min Act: -0.0002

Data Index: 1583106 (Python Code)

Max Activating Token Index: 8

Text #2

Max Range: 3.4508. Min Range: -3.4508

Max Act: 3.0576. Min Act: -0.0002

Data Index: 1473792 (Python Code)

Max Activating Token Index: 849

Text #3

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.9275. Min Act: -0.0002

Data Index: 1472573 (Python Code)

Max Activating Token Index: 10

Text #4

Max Range: 3.4508. Min Range: -3.4508

Max Act: 3.0481. Min Act: -0.0002

Data Index: 1625735 (Python Code)

Max Activating Token Index: 83

Text #5

Max Range: 3.4508. Min Range: -3.4508

Max Act: 3.0542. Min Act: -0.0002

Data Index: 1590945 (Python Code)

Max Activating Token Index: 10

Text #6

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.9222. Min Act: -0.0002

Data Index: 1491547 (Python Code)

Max Activating Token Index: 15

Text #7

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.7605. Min Act: -0.0002

Data Index: 1502150 (Python Code)

Max Activating Token Index: 6

Text #8

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.7611. Min Act: -0.0002

Data Index: 1579659 (Python Code)

Max Activating Token Index: 7

Text #9

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.6105. Min Act: -0.0002

Data Index: 1620276 (Python Code)

Max Activating Token Index: 8

Text #10

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.8707. Min Act: -0.0002

Data Index: 1479273 (Python Code)

Max Activating Token Index: 17

Text #11

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.9441. Min Act: -0.0002

Data Index: 1475340 (Python Code)

Max Activating Token Index: 12

Text #12

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.8353. Min Act: -0.0002

Data Index: 1632408 (Python Code)

Max Activating Token Index: 5

Text #13

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.8353. Min Act: -0.0002

Data Index: 1362448 (Python Code)

Max Activating Token Index: 5

Text #14

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.8353. Min Act: -0.0002

Data Index: 1530585 (Python Code)

Max Activating Token Index: 5

Text #15

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.8353. Min Act: -0.0002

Data Index: 1594180 (Python Code)

Max Activating Token Index: 5

Text #16

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.8353. Min Act: -0.0002

Data Index: 1540652 (Python Code)

Max Activating Token Index: 5

Text #17

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.8353. Min Act: -0.0002

Data Index: 1484332 (Python Code)

Max Activating Token Index: 5

Text #18

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.8353. Min Act: -0.0002

Data Index: 1447750 (Python Code)

Max Activating Token Index: 5

Text #19

Max Range: 3.4508. Min Range: -3.4508

Max Act: 2.8232. Min Act: -0.0002

Data Index: 1583821 (Python Code)

Max Activating Token Index: 10