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

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

Neuron 3545 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0834. Min Act: -0.0000

Data Index: 1052433 (C4 (Web Text))

Max Activating Token Index: 654

Text #1

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0884. Min Act: -0.0000

Data Index: 234809 (C4 (Web Text))

Max Activating Token Index: 407

Text #2

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0868. Min Act: -0.0000

Data Index: 254973 (C4 (Web Text))

Max Activating Token Index: 809

Text #3

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0869. Min Act: -0.0000

Data Index: 208598 (C4 (Web Text))

Max Activating Token Index: 818

Text #4

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0826. Min Act: -0.0001

Data Index: 826821 (C4 (Web Text))

Max Activating Token Index: 137

Text #5

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0779. Min Act: -0.0001

Data Index: 1352758 (C4 (Web Text))

Max Activating Token Index: 755

Text #6

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0693. Min Act: -0.0000

Data Index: 113775 (C4 (Web Text))

Max Activating Token Index: 465

Text #7

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0712. Min Act: -0.0000

Data Index: 892293 (C4 (Web Text))

Max Activating Token Index: 847

Text #8

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0717. Min Act: -0.0000

Data Index: 1096910 (C4 (Web Text))

Max Activating Token Index: 235

Text #9

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0656. Min Act: -0.0000

Data Index: 853541 (C4 (Web Text))

Max Activating Token Index: 829

Text #10

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0620. Min Act: -0.0001

Data Index: 1633014 (Python Code)

Max Activating Token Index: 413

Text #11

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0661. Min Act: -0.0000

Data Index: 251938 (C4 (Web Text))

Max Activating Token Index: 650

Text #12

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0637. Min Act: -0.0000

Data Index: 888963 (C4 (Web Text))

Max Activating Token Index: 229

Text #13

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0600. Min Act: -0.0000

Data Index: 874347 (C4 (Web Text))

Max Activating Token Index: 337

Text #14

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0628. Min Act: -0.0001

Data Index: 1123984 (C4 (Web Text))

Max Activating Token Index: 539

Text #15

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0648. Min Act: -0.0001

Data Index: 654959 (C4 (Web Text))

Max Activating Token Index: 799

Text #16

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0642. Min Act: -0.0000

Data Index: 384413 (C4 (Web Text))

Max Activating Token Index: 503

Text #17

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0631. Min Act: -0.0001

Data Index: 1342581 (C4 (Web Text))

Max Activating Token Index: 251

Text #18

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0646. Min Act: -0.0000

Data Index: 25664 (C4 (Web Text))

Max Activating Token Index: 974

Text #19

Max Range: 0.0884. Min Range: -0.0884

Max Act: 0.0608. Min Act: -0.0001

Data Index: 1584595 (Python Code)

Max Activating Token Index: 565