Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 5819 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0885. Min Act: -0.0001

Data Index: 1605360 (Python Code)

Max Activating Token Index: 43

Text #1

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0890. Min Act: -0.0001

Data Index: 699431 (C4 (Web Text))

Max Activating Token Index: 12

Text #2

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0831. Min Act: -0.0001

Data Index: 212487 (C4 (Web Text))

Max Activating Token Index: 376

Text #3

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0746. Min Act: -0.0001

Data Index: 1436528 (Python Code)

Max Activating Token Index: 19

Text #4

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0811. Min Act: -0.0001

Data Index: 832779 (C4 (Web Text))

Max Activating Token Index: 25

Text #5

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0843. Min Act: -0.0001

Data Index: 277832 (C4 (Web Text))

Max Activating Token Index: 253

Text #6

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0796. Min Act: -0.0001

Data Index: 785321 (C4 (Web Text))

Max Activating Token Index: 914

Text #7

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0820. Min Act: -0.0001

Data Index: 1289499 (C4 (Web Text))

Max Activating Token Index: 61

Text #8

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0765. Min Act: -0.0001

Data Index: 744451 (C4 (Web Text))

Max Activating Token Index: 572

Text #9

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0795. Min Act: -0.0001

Data Index: 211228 (C4 (Web Text))

Max Activating Token Index: 111

Text #10

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0700. Min Act: -0.0001

Data Index: 944868 (C4 (Web Text))

Max Activating Token Index: 231

Text #11

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0802. Min Act: -0.0001

Data Index: 773018 (C4 (Web Text))

Max Activating Token Index: 130

Text #12

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0761. Min Act: -0.0001

Data Index: 805605 (C4 (Web Text))

Max Activating Token Index: 948

Text #13

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0761. Min Act: -0.0001

Data Index: 924690 (C4 (Web Text))

Max Activating Token Index: 330

Text #14

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0798. Min Act: -0.0001

Data Index: 1102635 (C4 (Web Text))

Max Activating Token Index: 141

Text #15

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0756. Min Act: -0.0001

Data Index: 10240 (C4 (Web Text))

Max Activating Token Index: 515

Text #16

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0797. Min Act: -0.0001

Data Index: 360378 (C4 (Web Text))

Max Activating Token Index: 461

Text #17

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0810. Min Act: -0.0001

Data Index: 1454600 (Python Code)

Max Activating Token Index: 241

Text #18

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0781. Min Act: -0.0001

Data Index: 340209 (C4 (Web Text))

Max Activating Token Index: 75

Text #19

Max Range: 0.0890. Min Range: -0.0890

Max Act: 0.0826. Min Act: -0.0001

Data Index: 1399513 (Python Code)

Max Activating Token Index: 719