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

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

Neuron 512 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: 0.0829. Min Range: -0.0829

Max Act: 0.0829. Min Act: -0.0002

Data Index: 493480 (C4 (Web Text))

Max Activating Token Index: 710

Text #1

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0804. Min Act: -0.0003

Data Index: 459990 (C4 (Web Text))

Max Activating Token Index: 761

Text #2

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0805. Min Act: -0.0003

Data Index: 1066131 (C4 (Web Text))

Max Activating Token Index: 740

Text #3

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0790. Min Act: -0.0004

Data Index: 1004468 (C4 (Web Text))

Max Activating Token Index: 897

Text #4

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0800. Min Act: -0.0003

Data Index: 1145932 (C4 (Web Text))

Max Activating Token Index: 563

Text #5

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0786. Min Act: -0.0003

Data Index: 828086 (C4 (Web Text))

Max Activating Token Index: 779

Text #6

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0792. Min Act: -0.0003

Data Index: 862240 (C4 (Web Text))

Max Activating Token Index: 561

Text #7

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0759. Min Act: -0.0003

Data Index: 1038502 (C4 (Web Text))

Max Activating Token Index: 607

Text #8

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0764. Min Act: -0.0003

Data Index: 1286975 (C4 (Web Text))

Max Activating Token Index: 809

Text #9

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0775. Min Act: -0.0003

Data Index: 443351 (C4 (Web Text))

Max Activating Token Index: 695

Text #10

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0777. Min Act: -0.0003

Data Index: 997531 (C4 (Web Text))

Max Activating Token Index: 338

Text #11

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0755. Min Act: -0.0003

Data Index: 638367 (C4 (Web Text))

Max Activating Token Index: 740

Text #12

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0745. Min Act: -0.0003

Data Index: 1130937 (C4 (Web Text))

Max Activating Token Index: 917

Text #13

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0731. Min Act: -0.0003

Data Index: 835066 (C4 (Web Text))

Max Activating Token Index: 240

Text #14

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0756. Min Act: -0.0003

Data Index: 1226144 (C4 (Web Text))

Max Activating Token Index: 471

Text #15

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0723. Min Act: -0.0003

Data Index: 535329 (C4 (Web Text))

Max Activating Token Index: 882

Text #16

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0727. Min Act: -0.0003

Data Index: 1287772 (C4 (Web Text))

Max Activating Token Index: 369

Text #17

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0728. Min Act: -0.0003

Data Index: 647827 (C4 (Web Text))

Max Activating Token Index: 218

Text #18

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0773. Min Act: -0.0003

Data Index: 1080974 (C4 (Web Text))

Max Activating Token Index: 917

Text #19

Max Range: 0.0829. Min Range: -0.0829

Max Act: 0.0761. Min Act: -0.0003

Data Index: 448546 (C4 (Web Text))

Max Activating Token Index: 883