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

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

Neuron 391 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: 0.0880. Min Range: -0.0880

Max Act: 0.0880. Min Act: -0.0001

Data Index: 1218115 (C4 (Web Text))

Max Activating Token Index: 490

Text #1

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0752. Min Act: -0.0001

Data Index: 1332846 (C4 (Web Text))

Max Activating Token Index: 453

Text #2

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0672. Min Act: -0.0001

Data Index: 1353038 (C4 (Web Text))

Max Activating Token Index: 619

Text #3

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0672. Min Act: -0.0001

Data Index: 94523 (C4 (Web Text))

Max Activating Token Index: 36

Text #4

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0600. Min Act: -0.0001

Data Index: 1626547 (Python Code)

Max Activating Token Index: 70

Text #5

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0539. Min Act: -0.0001

Data Index: 552061 (C4 (Web Text))

Max Activating Token Index: 183

Text #6

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0399. Min Act: -0.0001

Data Index: 1290855 (C4 (Web Text))

Max Activating Token Index: 253

Text #7

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0582. Min Act: -0.0001

Data Index: 1278246 (C4 (Web Text))

Max Activating Token Index: 718

Text #8

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0603. Min Act: -0.0001

Data Index: 1194903 (C4 (Web Text))

Max Activating Token Index: 49

Text #9

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0623. Min Act: -0.0001

Data Index: 191378 (C4 (Web Text))

Max Activating Token Index: 264

Text #10

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0593. Min Act: -0.0001

Data Index: 1121123 (C4 (Web Text))

Max Activating Token Index: 42

Text #11

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0638. Min Act: -0.0001

Data Index: 1055336 (C4 (Web Text))

Max Activating Token Index: 34

Text #12

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0537. Min Act: -0.0001

Data Index: 532480 (C4 (Web Text))

Max Activating Token Index: 617

Text #13

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0600. Min Act: -0.0001

Data Index: 1416546 (Python Code)

Max Activating Token Index: 53

Text #14

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0524. Min Act: -0.0001

Data Index: 802993 (C4 (Web Text))

Max Activating Token Index: 829

Text #15

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0549. Min Act: -0.0001

Data Index: 798496 (C4 (Web Text))

Max Activating Token Index: 55

Text #16

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0462. Min Act: -0.0001

Data Index: 1461039 (Python Code)

Max Activating Token Index: 321

Text #17

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0516. Min Act: -0.0001

Data Index: 53843 (C4 (Web Text))

Max Activating Token Index: 501

Text #18

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0608. Min Act: -0.0002

Data Index: 609776 (C4 (Web Text))

Max Activating Token Index: 871

Text #19

Max Range: 0.0880. Min Range: -0.0880

Max Act: 0.0525. Min Act: -0.0001

Data Index: 1524054 (Python Code)

Max Activating Token Index: 150