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

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

Neuron 3525 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.7438. Min Range: -3.7438

Max Act: 3.7438. Min Act: -0.0000

Data Index: 856096 (C4 (Web Text))

Max Activating Token Index: 405

Text #1

Max Range: 3.7438. Min Range: -3.7438

Max Act: 3.5529. Min Act: -0.0000

Data Index: 352273 (C4 (Web Text))

Max Activating Token Index: 977

Text #2

Max Range: 3.7438. Min Range: -3.7438

Max Act: 3.4833. Min Act: -0.0000

Data Index: 203628 (C4 (Web Text))

Max Activating Token Index: 49

Text #3

Max Range: 3.7438. Min Range: -3.7438

Max Act: 3.4259. Min Act: 0.0000

Data Index: 792518 (C4 (Web Text))

Max Activating Token Index: 757

Text #4

Max Range: 3.7438. Min Range: -3.7438

Max Act: 3.3478. Min Act: -0.0000

Data Index: 997808 (C4 (Web Text))

Max Activating Token Index: 391

Text #5

Max Range: 3.7438. Min Range: -3.7438

Max Act: 3.0733. Min Act: -0.0000

Data Index: 613058 (C4 (Web Text))

Max Activating Token Index: 728

Text #6

Max Range: 3.7438. Min Range: -3.7438

Max Act: 3.1957. Min Act: -0.0000

Data Index: 252439 (C4 (Web Text))

Max Activating Token Index: 965

Text #7

Max Range: 3.7438. Min Range: -3.7438

Max Act: 3.1846. Min Act: -0.0000

Data Index: 117306 (C4 (Web Text))

Max Activating Token Index: 713

Text #8

Max Range: 3.7438. Min Range: -3.7438

Max Act: 3.1371. Min Act: -0.0000

Data Index: 1633555 (Python Code)

Max Activating Token Index: 763

Text #9

Max Range: 3.7438. Min Range: -3.7438

Max Act: 2.8493. Min Act: -0.0000

Data Index: 1422797 (Python Code)

Max Activating Token Index: 409

Text #10

Max Range: 3.7438. Min Range: -3.7438

Max Act: 2.7661. Min Act: -0.0000

Data Index: 947828 (C4 (Web Text))

Max Activating Token Index: 875

Text #11

Max Range: 3.7438. Min Range: -3.7438

Max Act: 2.7464. Min Act: -0.0000

Data Index: 1496951 (Python Code)

Max Activating Token Index: 653

Text #12

Max Range: 3.7438. Min Range: -3.7438

Max Act: 2.7398. Min Act: -0.0000

Data Index: 1628972 (Python Code)

Max Activating Token Index: 984

Text #13

Max Range: 3.7438. Min Range: -3.7438

Max Act: 3.2298. Min Act: -0.0000

Data Index: 1601665 (Python Code)

Max Activating Token Index: 218

Text #14

Max Range: 3.7438. Min Range: -3.7438

Max Act: 2.8226. Min Act: -0.0000

Data Index: 1394677 (Python Code)

Max Activating Token Index: 987

Text #15

Max Range: 3.7438. Min Range: -3.7438

Max Act: 2.6563. Min Act: -0.0000

Data Index: 947502 (C4 (Web Text))

Max Activating Token Index: 332

Text #16

Max Range: 3.7438. Min Range: -3.7438

Max Act: 2.4025. Min Act: -0.0000

Data Index: 1417742 (Python Code)

Max Activating Token Index: 271

Text #17

Max Range: 3.7438. Min Range: -3.7438

Max Act: 2.6335. Min Act: -0.0000

Data Index: 1632744 (Python Code)

Max Activating Token Index: 875

Text #18

Max Range: 3.7438. Min Range: -3.7438

Max Act: 2.9203. Min Act: -0.0000

Data Index: 101576 (C4 (Web Text))

Max Activating Token Index: 795

Text #19

Max Range: 3.7438. Min Range: -3.7438

Max Act: 2.7868. Min Act: -0.0000

Data Index: 823387 (C4 (Web Text))

Max Activating Token Index: 883