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

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

Neuron 4579 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: 0.7863. Min Range: -0.7863

Max Act: 0.7678. Min Act: -0.0000

Data Index: 1643353 (Python Code)

Max Activating Token Index: 28

Text #1

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.7863. Min Act: -0.0000

Data Index: 1451381 (Python Code)

Max Activating Token Index: 159

Text #2

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.7342. Min Act: -0.0000

Data Index: 1586723 (Python Code)

Max Activating Token Index: 364

Text #3

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.7451. Min Act: -0.0000

Data Index: 1414349 (Python Code)

Max Activating Token Index: 362

Text #4

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.7360. Min Act: -0.0000

Data Index: 1071206 (C4 (Web Text))

Max Activating Token Index: 81

Text #5

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.6863. Min Act: -0.0000

Data Index: 1540435 (Python Code)

Max Activating Token Index: 35

Text #6

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.7118. Min Act: -0.0000

Data Index: 1581073 (Python Code)

Max Activating Token Index: 398

Text #7

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.7271. Min Act: -0.0000

Data Index: 2864 (C4 (Web Text))

Max Activating Token Index: 638

Text #8

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.7692. Min Act: -0.0000

Data Index: 1063790 (C4 (Web Text))

Max Activating Token Index: 50

Text #9

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.7313. Min Act: -0.0000

Data Index: 568241 (C4 (Web Text))

Max Activating Token Index: 307

Text #10

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.7101. Min Act: -0.0000

Data Index: 534648 (C4 (Web Text))

Max Activating Token Index: 501

Text #11

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.6840. Min Act: -0.0000

Data Index: 739238 (C4 (Web Text))

Max Activating Token Index: 673

Text #12

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.7232. Min Act: -0.0000

Data Index: 1291590 (C4 (Web Text))

Max Activating Token Index: 147

Text #13

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.6955. Min Act: -0.0000

Data Index: 1018659 (C4 (Web Text))

Max Activating Token Index: 741

Text #14

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.6950. Min Act: -0.0000

Data Index: 1466545 (Python Code)

Max Activating Token Index: 48

Text #15

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.7037. Min Act: -0.0000

Data Index: 444296 (C4 (Web Text))

Max Activating Token Index: 193

Text #16

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.7033. Min Act: -0.0000

Data Index: 1135556 (C4 (Web Text))

Max Activating Token Index: 60

Text #17

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.6308. Min Act: -0.0000

Data Index: 1447631 (Python Code)

Max Activating Token Index: 222

Text #18

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.6330. Min Act: -0.0000

Data Index: 462019 (C4 (Web Text))

Max Activating Token Index: 303

Text #19

Max Range: 0.7863. Min Range: -0.7863

Max Act: 0.6355. Min Act: -0.0000

Data Index: 982334 (C4 (Web Text))

Max Activating Token Index: 565