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

Dataset: The Pile

Neuron 3420 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 8.9664. Min Range: -8.9664

Max Act: 7.2344. Min Act: -0.0000

Data Index: 523174 (The Pile)

Max Activating Token Index: 866

Text #1

Max Range: 8.9664. Min Range: -8.9664

Max Act: 7.4926. Min Act: -0.0000

Data Index: 1631762 (The Pile)

Max Activating Token Index: 87

Text #2

Max Range: 8.9664. Min Range: -8.9664

Max Act: 8.9132. Min Act: -0.0000

Data Index: 390782 (The Pile)

Max Activating Token Index: 224

Text #3

Max Range: 8.9664. Min Range: -8.9664

Max Act: 8.7958. Min Act: -0.0000

Data Index: 189774 (The Pile)

Max Activating Token Index: 794

Text #4

Max Range: 8.9664. Min Range: -8.9664

Max Act: 6.9051. Min Act: -0.0000

Data Index: 732897 (The Pile)

Max Activating Token Index: 906

Text #5

Max Range: 8.9664. Min Range: -8.9664

Max Act: 8.2458. Min Act: -0.0000

Data Index: 519445 (The Pile)

Max Activating Token Index: 398

Text #6

Max Range: 8.9664. Min Range: -8.9664

Max Act: 8.0670. Min Act: -0.0000

Data Index: 1849651 (The Pile)

Max Activating Token Index: 319

Text #7

Max Range: 8.9664. Min Range: -8.9664

Max Act: 8.1869. Min Act: -0.0000

Data Index: 416969 (The Pile)

Max Activating Token Index: 70

Text #8

Max Range: 8.9664. Min Range: -8.9664

Max Act: 7.2667. Min Act: -0.0000

Data Index: 492715 (The Pile)

Max Activating Token Index: 614

Text #9

Max Range: 8.9664. Min Range: -8.9664

Max Act: 8.3106. Min Act: -0.0000

Data Index: 19106 (The Pile)

Max Activating Token Index: 220

Text #10

Max Range: 8.9664. Min Range: -8.9664

Max Act: 8.9664. Min Act: -0.0000

Data Index: 1123953 (The Pile)

Max Activating Token Index: 181

Text #11

Max Range: 8.9664. Min Range: -8.9664

Max Act: 8.0757. Min Act: -0.0000

Data Index: 1672477 (The Pile)

Max Activating Token Index: 303

Text #12

Max Range: 8.9664. Min Range: -8.9664

Max Act: 8.2228. Min Act: -0.0000

Data Index: 1038229 (The Pile)

Max Activating Token Index: 151

Text #13

Max Range: 8.9664. Min Range: -8.9664

Max Act: 7.9671. Min Act: -0.0000

Data Index: 1938094 (The Pile)

Max Activating Token Index: 147

Text #14

Max Range: 8.9664. Min Range: -8.9664

Max Act: 7.8999. Min Act: -0.0000

Data Index: 166223 (The Pile)

Max Activating Token Index: 136

Text #15

Max Range: 8.9664. Min Range: -8.9664

Max Act: 7.7399. Min Act: -0.0000

Data Index: 593784 (The Pile)

Max Activating Token Index: 190

Text #16

Max Range: 8.9664. Min Range: -8.9664

Max Act: 7.6636. Min Act: -0.0000

Data Index: 1790887 (The Pile)

Max Activating Token Index: 323

Text #17

Max Range: 8.9664. Min Range: -8.9664

Max Act: 7.4490. Min Act: -0.0000

Data Index: 366503 (The Pile)

Max Activating Token Index: 827

Text #18

Max Range: 8.9664. Min Range: -8.9664

Max Act: 8.0498. Min Act: -0.0000

Data Index: 1510415 (The Pile)

Max Activating Token Index: 258

Text #19

Max Range: 8.9664. Min Range: -8.9664

Max Act: 7.8158. Min Act: -0.0000

Data Index: 59991 (The Pile)

Max Activating Token Index: 328