Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 3642 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 1.1522. Min Range: -1.1522

Max Act: 1.1177. Min Act: -0.0001

Data Index: 984679 (C4 (Web Text))

Max Activating Token Index: 335

Text #1

Max Range: 1.1522. Min Range: -1.1522

Max Act: 1.1522. Min Act: -0.0001

Data Index: 211545 (C4 (Web Text))

Max Activating Token Index: 300

Text #2

Max Range: 1.1522. Min Range: -1.1522

Max Act: 1.1333. Min Act: -0.0001

Data Index: 281411 (C4 (Web Text))

Max Activating Token Index: 478

Text #3

Max Range: 1.1522. Min Range: -1.1522

Max Act: 1.1290. Min Act: -0.0001

Data Index: 1626410 (Python Code)

Max Activating Token Index: 355

Text #4

Max Range: 1.1522. Min Range: -1.1522

Max Act: 1.1305. Min Act: -0.0001

Data Index: 1492869 (Python Code)

Max Activating Token Index: 930

Text #5

Max Range: 1.1522. Min Range: -1.1522

Max Act: 1.0225. Min Act: -0.0001

Data Index: 1028218 (C4 (Web Text))

Max Activating Token Index: 383

Text #6

Max Range: 1.1522. Min Range: -1.1522

Max Act: 0.9986. Min Act: -0.0001

Data Index: 1438319 (Python Code)

Max Activating Token Index: 189

Text #7

Max Range: 1.1522. Min Range: -1.1522

Max Act: 0.9664. Min Act: -0.0001

Data Index: 965059 (C4 (Web Text))

Max Activating Token Index: 875

Text #8

Max Range: 1.1522. Min Range: -1.1522

Max Act: 1.0479. Min Act: -0.0001

Data Index: 1447808 (Python Code)

Max Activating Token Index: 788

Text #9

Max Range: 1.1522. Min Range: -1.1522

Max Act: 1.0263. Min Act: -0.0001

Data Index: 422145 (C4 (Web Text))

Max Activating Token Index: 716

Text #10

Max Range: 1.1522. Min Range: -1.1522

Max Act: 0.9805. Min Act: -0.0001

Data Index: 1268791 (C4 (Web Text))

Max Activating Token Index: 289

Text #11

Max Range: 1.1522. Min Range: -1.1522

Max Act: 0.9300. Min Act: -0.0001

Data Index: 1399544 (Python Code)

Max Activating Token Index: 465

Text #12

Max Range: 1.1522. Min Range: -1.1522

Max Act: 0.9046. Min Act: -0.0001

Data Index: 1547848 (Python Code)

Max Activating Token Index: 991

Text #13

Max Range: 1.1522. Min Range: -1.1522

Max Act: 0.9729. Min Act: -0.0001

Data Index: 799709 (C4 (Web Text))

Max Activating Token Index: 128

Text #14

Max Range: 1.1522. Min Range: -1.1522

Max Act: 0.8915. Min Act: -0.0001

Data Index: 1092446 (C4 (Web Text))

Max Activating Token Index: 361

Text #15

Max Range: 1.1522. Min Range: -1.1522

Max Act: 0.8967. Min Act: -0.0001

Data Index: 360344 (C4 (Web Text))

Max Activating Token Index: 576

Text #16

Max Range: 1.1522. Min Range: -1.1522

Max Act: 0.8956. Min Act: -0.0001

Data Index: 131791 (C4 (Web Text))

Max Activating Token Index: 445

Text #17

Max Range: 1.1522. Min Range: -1.1522

Max Act: 0.9753. Min Act: -0.0001

Data Index: 106380 (C4 (Web Text))

Max Activating Token Index: 978

Text #18

Max Range: 1.1522. Min Range: -1.1522

Max Act: 0.9434. Min Act: -0.0001

Data Index: 1603854 (Python Code)

Max Activating Token Index: 137

Text #19

Max Range: 1.1522. Min Range: -1.1522

Max Act: 0.9671. Min Act: -0.0001

Data Index: 190085 (C4 (Web Text))

Max Activating Token Index: 977