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

Dataset: The Pile

Neuron 4023 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 9.3966. Min Range: -9.3966

Max Act: 9.3966. Min Act: -0.0000

Data Index: 1068949 (The Pile)

Max Activating Token Index: 350

Text #1

Max Range: 9.3966. Min Range: -9.3966

Max Act: 9.0406. Min Act: -0.0000

Data Index: 113733 (The Pile)

Max Activating Token Index: 862

Text #2

Max Range: 9.3966. Min Range: -9.3966

Max Act: 8.9264. Min Act: -0.0000

Data Index: 26517 (The Pile)

Max Activating Token Index: 920

Text #3

Max Range: 9.3966. Min Range: -9.3966

Max Act: 8.4301. Min Act: -0.0000

Data Index: 1974039 (The Pile)

Max Activating Token Index: 42

Text #4

Max Range: 9.3966. Min Range: -9.3966

Max Act: 8.4573. Min Act: -0.0000

Data Index: 1480164 (The Pile)

Max Activating Token Index: 30

Text #5

Max Range: 9.3966. Min Range: -9.3966

Max Act: 8.7779. Min Act: -0.0000

Data Index: 1240060 (The Pile)

Max Activating Token Index: 40

Text #6

Max Range: 9.3966. Min Range: -9.3966

Max Act: 8.2214. Min Act: -0.0000

Data Index: 1592505 (The Pile)

Max Activating Token Index: 215

Text #7

Max Range: 9.3966. Min Range: -9.3966

Max Act: 7.9580. Min Act: -0.0000

Data Index: 463149 (The Pile)

Max Activating Token Index: 183

Text #8

Max Range: 9.3966. Min Range: -9.3966

Max Act: 8.3197. Min Act: -0.0000

Data Index: 906195 (The Pile)

Max Activating Token Index: 392

Text #9

Max Range: 9.3966. Min Range: -9.3966

Max Act: 8.5794. Min Act: -0.0000

Data Index: 109114 (The Pile)

Max Activating Token Index: 527

Text #10

Max Range: 9.3966. Min Range: -9.3966

Max Act: 7.8148. Min Act: -0.0000

Data Index: 889962 (The Pile)

Max Activating Token Index: 146

Text #11

Max Range: 9.3966. Min Range: -9.3966

Max Act: 7.9162. Min Act: -0.0000

Data Index: 1167544 (The Pile)

Max Activating Token Index: 231

Text #12

Max Range: 9.3966. Min Range: -9.3966

Max Act: 8.0413. Min Act: -0.0000

Data Index: 1842539 (The Pile)

Max Activating Token Index: 1014

Text #13

Max Range: 9.3966. Min Range: -9.3966

Max Act: 8.0119. Min Act: -0.0000

Data Index: 712169 (The Pile)

Max Activating Token Index: 868

Text #14

Max Range: 9.3966. Min Range: -9.3966

Max Act: 8.2986. Min Act: -0.0000

Data Index: 1722392 (The Pile)

Max Activating Token Index: 82

Text #15

Max Range: 9.3966. Min Range: -9.3966

Max Act: 7.6657. Min Act: -0.0000

Data Index: 381665 (The Pile)

Max Activating Token Index: 663

Text #16

Max Range: 9.3966. Min Range: -9.3966

Max Act: 7.8934. Min Act: -0.0000

Data Index: 794872 (The Pile)

Max Activating Token Index: 195

Text #17

Max Range: 9.3966. Min Range: -9.3966

Max Act: 7.8224. Min Act: -0.0000

Data Index: 890355 (The Pile)

Max Activating Token Index: 7

Text #18

Max Range: 9.3966. Min Range: -9.3966

Max Act: 7.7812. Min Act: -0.0000

Data Index: 563254 (The Pile)

Max Activating Token Index: 221

Text #19

Max Range: 9.3966. Min Range: -9.3966

Max Act: 7.9766. Min Act: -0.0000

Data Index: 1278866 (The Pile)

Max Activating Token Index: 58