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

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

Neuron 1051 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.9120. Min Act: -0.0002

Data Index: 1540579 (Python Code)

Max Activating Token Index: 816

Text #1

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.9722. Min Act: -0.0002

Data Index: 470059 (C4 (Web Text))

Max Activating Token Index: 676

Text #2

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.8780. Min Act: -0.0002

Data Index: 1509768 (Python Code)

Max Activating Token Index: 8

Text #3

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.7338. Min Act: -0.0002

Data Index: 1248773 (C4 (Web Text))

Max Activating Token Index: 188

Text #4

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.8683. Min Act: -0.0002

Data Index: 759884 (C4 (Web Text))

Max Activating Token Index: 161

Text #5

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.7854. Min Act: -0.0002

Data Index: 879150 (C4 (Web Text))

Max Activating Token Index: 658

Text #6

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.6912. Min Act: -0.0002

Data Index: 905162 (C4 (Web Text))

Max Activating Token Index: 34

Text #7

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.7745. Min Act: -0.0002

Data Index: 471499 (C4 (Web Text))

Max Activating Token Index: 570

Text #8

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.7036. Min Act: -0.0002

Data Index: 164644 (C4 (Web Text))

Max Activating Token Index: 813

Text #9

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.7545. Min Act: -0.0002

Data Index: 817708 (C4 (Web Text))

Max Activating Token Index: 244

Text #10

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.7401. Min Act: -0.0002

Data Index: 551560 (C4 (Web Text))

Max Activating Token Index: 21

Text #11

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.6441. Min Act: -0.0002

Data Index: 1404989 (Python Code)

Max Activating Token Index: 736

Text #12

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.7347. Min Act: -0.0002

Data Index: 1655859 (Python Code)

Max Activating Token Index: 1002

Text #13

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.7446. Min Act: -0.0002

Data Index: 489503 (C4 (Web Text))

Max Activating Token Index: 963

Text #14

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.7602. Min Act: -0.0002

Data Index: 1525541 (Python Code)

Max Activating Token Index: 61

Text #15

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.5719. Min Act: -0.0002

Data Index: 1075734 (C4 (Web Text))

Max Activating Token Index: 172

Text #16

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.8708. Min Act: -0.0002

Data Index: 267320 (C4 (Web Text))

Max Activating Token Index: 889

Text #17

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.8177. Min Act: -0.0002

Data Index: 1207048 (C4 (Web Text))

Max Activating Token Index: 32

Text #18

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.6344. Min Act: -0.0002

Data Index: 825302 (C4 (Web Text))

Max Activating Token Index: 839

Text #19

Max Range: 0.9722. Min Range: -0.9722

Max Act: 0.7017. Min Act: -0.0002

Data Index: 777439 (C4 (Web Text))

Max Activating Token Index: 935