Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 1097 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0621. Min Act: -0.0003

Data Index: 92077 (C4 (Web Text))

Max Activating Token Index: 347

Text #1

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0551. Min Act: -0.0004

Data Index: 656755 (C4 (Web Text))

Max Activating Token Index: 932

Text #2

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0539. Min Act: -0.0003

Data Index: 649410 (C4 (Web Text))

Max Activating Token Index: 37

Text #3

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0544. Min Act: -0.0003

Data Index: 960318 (C4 (Web Text))

Max Activating Token Index: 644

Text #4

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0518. Min Act: -0.0003

Data Index: 1595779 (Python Code)

Max Activating Token Index: 390

Text #5

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0514. Min Act: -0.0003

Data Index: 26070 (C4 (Web Text))

Max Activating Token Index: 938

Text #6

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0531. Min Act: -0.0002

Data Index: 164264 (C4 (Web Text))

Max Activating Token Index: 41

Text #7

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0505. Min Act: -0.0003

Data Index: 640391 (C4 (Web Text))

Max Activating Token Index: 824

Text #8

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0508. Min Act: -0.0003

Data Index: 837646 (C4 (Web Text))

Max Activating Token Index: 357

Text #9

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0506. Min Act: -0.0003

Data Index: 234546 (C4 (Web Text))

Max Activating Token Index: 953

Text #10

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0498. Min Act: -0.0003

Data Index: 619227 (C4 (Web Text))

Max Activating Token Index: 736

Text #11

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0507. Min Act: -0.0003

Data Index: 715986 (C4 (Web Text))

Max Activating Token Index: 818

Text #12

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0501. Min Act: -0.0003

Data Index: 650954 (C4 (Web Text))

Max Activating Token Index: 420

Text #13

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0503. Min Act: -0.0003

Data Index: 953159 (C4 (Web Text))

Max Activating Token Index: 84

Text #14

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0497. Min Act: -0.0002

Data Index: 1068517 (C4 (Web Text))

Max Activating Token Index: 816

Text #15

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0501. Min Act: -0.0003

Data Index: 1641023 (Python Code)

Max Activating Token Index: 331

Text #16

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0481. Min Act: -0.0002

Data Index: 836064 (C4 (Web Text))

Max Activating Token Index: 722

Text #17

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0504. Min Act: -0.0002

Data Index: 1023577 (C4 (Web Text))

Max Activating Token Index: 935

Text #18

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0486. Min Act: -0.0003

Data Index: 1238562 (C4 (Web Text))

Max Activating Token Index: 750

Text #19

Max Range: 0.0621. Min Range: -0.0621

Max Act: 0.0483. Min Act: -0.0003

Data Index: 1172131 (C4 (Web Text))

Max Activating Token Index: 466