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

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

Neuron 348 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.5817. Min Act: -0.0000

Data Index: 820077 (C4 (Web Text))

Max Activating Token Index: 22

Text #1

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.5128. Min Act: -0.0000

Data Index: 483255 (C4 (Web Text))

Max Activating Token Index: 791

Text #2

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.5028. Min Act: -0.0000

Data Index: 1036940 (C4 (Web Text))

Max Activating Token Index: 1008

Text #3

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.5590. Min Act: -0.0000

Data Index: 890921 (C4 (Web Text))

Max Activating Token Index: 141

Text #4

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.3931. Min Act: -0.0000

Data Index: 167753 (C4 (Web Text))

Max Activating Token Index: 182

Text #5

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.4478. Min Act: -0.0000

Data Index: 1271623 (C4 (Web Text))

Max Activating Token Index: 677

Text #6

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.4411. Min Act: -0.0000

Data Index: 1403170 (Python Code)

Max Activating Token Index: 30

Text #7

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.4138. Min Act: -0.0000

Data Index: 70525 (C4 (Web Text))

Max Activating Token Index: 473

Text #8

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.4052. Min Act: -0.0000

Data Index: 612402 (C4 (Web Text))

Max Activating Token Index: 242

Text #9

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.3851. Min Act: -0.0000

Data Index: 785508 (C4 (Web Text))

Max Activating Token Index: 581

Text #10

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.3745. Min Act: -0.0000

Data Index: 374780 (C4 (Web Text))

Max Activating Token Index: 638

Text #11

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.3782. Min Act: -0.0000

Data Index: 784173 (C4 (Web Text))

Max Activating Token Index: 729

Text #12

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.3691. Min Act: -0.0000

Data Index: 1603354 (Python Code)

Max Activating Token Index: 544

Text #13

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.3611. Min Act: -0.0000

Data Index: 860850 (C4 (Web Text))

Max Activating Token Index: 112

Text #14

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.3473. Min Act: -0.0000

Data Index: 1134062 (C4 (Web Text))

Max Activating Token Index: 900

Text #15

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.3272. Min Act: -0.0000

Data Index: 1007 (C4 (Web Text))

Max Activating Token Index: 801

Text #16

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.3092. Min Act: -0.0000

Data Index: 1142089 (C4 (Web Text))

Max Activating Token Index: 175

Text #17

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.2684. Min Act: -0.0000

Data Index: 529047 (C4 (Web Text))

Max Activating Token Index: 272

Text #18

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.1792. Min Act: -0.0000

Data Index: 1183767 (C4 (Web Text))

Max Activating Token Index: 159

Text #19

Max Range: 1.5817. Min Range: -1.5817

Max Act: 1.2545. Min Act: -0.0000

Data Index: 822674 (C4 (Web Text))

Max Activating Token Index: 976