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

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

Neuron 2839 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: 0.1590. Min Range: -0.1590

Max Act: 0.1590. Min Act: -0.0000

Data Index: 150478 (C4 (Web Text))

Max Activating Token Index: 775

Text #1

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.1326. Min Act: -0.0000

Data Index: 1301412 (C4 (Web Text))

Max Activating Token Index: 518

Text #2

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.1318. Min Act: -0.0000

Data Index: 285280 (C4 (Web Text))

Max Activating Token Index: 389

Text #3

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.1120. Min Act: -0.0001

Data Index: 874422 (C4 (Web Text))

Max Activating Token Index: 608

Text #4

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.1130. Min Act: -0.0000

Data Index: 965445 (C4 (Web Text))

Max Activating Token Index: 411

Text #5

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.1043. Min Act: -0.0000

Data Index: 106456 (C4 (Web Text))

Max Activating Token Index: 481

Text #6

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.1023. Min Act: -0.0000

Data Index: 173917 (C4 (Web Text))

Max Activating Token Index: 501

Text #7

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.1011. Min Act: -0.0000

Data Index: 846930 (C4 (Web Text))

Max Activating Token Index: 672

Text #8

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.0987. Min Act: -0.0000

Data Index: 1650345 (Python Code)

Max Activating Token Index: 993

Text #9

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.0990. Min Act: -0.0001

Data Index: 152427 (C4 (Web Text))

Max Activating Token Index: 372

Text #10

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.1050. Min Act: -0.0000

Data Index: 1425015 (Python Code)

Max Activating Token Index: 585

Text #11

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.0938. Min Act: -0.0001

Data Index: 753323 (C4 (Web Text))

Max Activating Token Index: 558

Text #12

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.0862. Min Act: -0.0000

Data Index: 1192637 (C4 (Web Text))

Max Activating Token Index: 840

Text #13

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.0878. Min Act: -0.0000

Data Index: 413893 (C4 (Web Text))

Max Activating Token Index: 375

Text #14

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.0962. Min Act: -0.0000

Data Index: 980670 (C4 (Web Text))

Max Activating Token Index: 242

Text #15

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.0928. Min Act: -0.0001

Data Index: 498920 (C4 (Web Text))

Max Activating Token Index: 480

Text #16

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.0889. Min Act: -0.0000

Data Index: 859122 (C4 (Web Text))

Max Activating Token Index: 550

Text #17

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.0883. Min Act: -0.0001

Data Index: 139741 (C4 (Web Text))

Max Activating Token Index: 400

Text #18

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.0919. Min Act: -0.0000

Data Index: 84617 (C4 (Web Text))

Max Activating Token Index: 156

Text #19

Max Range: 0.1590. Min Range: -0.1590

Max Act: 0.0787. Min Act: -0.0000

Data Index: 1571940 (Python Code)

Max Activating Token Index: 982