Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 3593 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 10.6645. Min Range: -10.6645

Max Act: 8.8950. Min Act: -0.0000

Data Index: 1580345 (Python Code)

Max Activating Token Index: 12

Text #1

Max Range: 10.6645. Min Range: -10.6645

Max Act: 10.6645. Min Act: -0.0001

Data Index: 1134411 (C4 (Web Text))

Max Activating Token Index: 432

Text #2

Max Range: 10.6645. Min Range: -10.6645

Max Act: 9.8756. Min Act: -0.0000

Data Index: 1461681 (Python Code)

Max Activating Token Index: 705

Text #3

Max Range: 10.6645. Min Range: -10.6645

Max Act: 9.3081. Min Act: -0.0000

Data Index: 640821 (C4 (Web Text))

Max Activating Token Index: 10

Text #4

Max Range: 10.6645. Min Range: -10.6645

Max Act: 8.9551. Min Act: -0.0000

Data Index: 651679 (C4 (Web Text))

Max Activating Token Index: 22

Text #5

Max Range: 10.6645. Min Range: -10.6645

Max Act: 8.5327. Min Act: -0.0001

Data Index: 460643 (C4 (Web Text))

Max Activating Token Index: 115

Text #6

Max Range: 10.6645. Min Range: -10.6645

Max Act: 8.2068. Min Act: -0.0000

Data Index: 1437024 (Python Code)

Max Activating Token Index: 20

Text #7

Max Range: 10.6645. Min Range: -10.6645

Max Act: 4.5376. Min Act: -0.0000

Data Index: 849405 (C4 (Web Text))

Max Activating Token Index: 32

Text #8

Max Range: 10.6645. Min Range: -10.6645

Max Act: 4.9138. Min Act: -0.0000

Data Index: 16247 (C4 (Web Text))

Max Activating Token Index: 211

Text #9

Max Range: 10.6645. Min Range: -10.6645

Max Act: 0.3642. Min Act: -0.0000

Data Index: 175808 (C4 (Web Text))

Max Activating Token Index: 138

Text #10

Max Range: 10.6645. Min Range: -10.6645

Max Act: 6.0931. Min Act: -0.0000

Data Index: 292510 (C4 (Web Text))

Max Activating Token Index: 481

Text #11

Max Range: 10.6645. Min Range: -10.6645

Max Act: 6.8744. Min Act: -0.0001

Data Index: 90265 (C4 (Web Text))

Max Activating Token Index: 57

Text #12

Max Range: 10.6645. Min Range: -10.6645

Max Act: 5.9024. Min Act: 0.0000

Data Index: 161160 (C4 (Web Text))

Max Activating Token Index: 226

Text #13

Max Range: 10.6645. Min Range: -10.6645

Max Act: 6.0316. Min Act: 0.0001

Data Index: 348138 (C4 (Web Text))

Max Activating Token Index: 53

Text #14

Max Range: 10.6645. Min Range: -10.6645

Max Act: 3.9198. Min Act: -0.0000

Data Index: 1351432 (C4 (Web Text))

Max Activating Token Index: 576

Text #15

Max Range: 10.6645. Min Range: -10.6645

Max Act: 3.7693. Min Act: -0.0000

Data Index: 408136 (C4 (Web Text))

Max Activating Token Index: 149

Text #16

Max Range: 10.6645. Min Range: -10.6645

Max Act: 2.5301. Min Act: -0.0000

Data Index: 550582 (C4 (Web Text))

Max Activating Token Index: 178

Text #17

Max Range: 10.6645. Min Range: -10.6645

Max Act: 3.6156. Min Act: -0.0001

Data Index: 794965 (C4 (Web Text))

Max Activating Token Index: 146

Text #18

Max Range: 10.6645. Min Range: -10.6645

Max Act: 1.8760. Min Act: -0.0000

Data Index: 945726 (C4 (Web Text))

Max Activating Token Index: 19

Text #19

Max Range: 10.6645. Min Range: -10.6645

Max Act: 0.1855. Min Act: -0.0000

Data Index: 279047 (C4 (Web Text))

Max Activating Token Index: 71