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

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

Neuron 3910 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.3895. Min Act: -0.0001

Data Index: 1419466 (Python Code)

Max Activating Token Index: 299

Text #1

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.3287. Min Act: -0.0001

Data Index: 922254 (C4 (Web Text))

Max Activating Token Index: 104

Text #2

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.3148. Min Act: -0.0001

Data Index: 1061570 (C4 (Web Text))

Max Activating Token Index: 337

Text #3

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.3134. Min Act: -0.0001

Data Index: 971980 (C4 (Web Text))

Max Activating Token Index: 98

Text #4

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.2760. Min Act: -0.0001

Data Index: 1099667 (C4 (Web Text))

Max Activating Token Index: 108

Text #5

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.3043. Min Act: -0.0001

Data Index: 1426049 (Python Code)

Max Activating Token Index: 981

Text #6

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.2940. Min Act: -0.0001

Data Index: 1419132 (Python Code)

Max Activating Token Index: 162

Text #7

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.2952. Min Act: -0.0001

Data Index: 958229 (C4 (Web Text))

Max Activating Token Index: 8

Text #8

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.3125. Min Act: -0.0001

Data Index: 1115527 (C4 (Web Text))

Max Activating Token Index: 303

Text #9

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.2827. Min Act: -0.0001

Data Index: 1458596 (Python Code)

Max Activating Token Index: 468

Text #10

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.2929. Min Act: -0.0001

Data Index: 964523 (C4 (Web Text))

Max Activating Token Index: 387

Text #11

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.3002. Min Act: -0.0001

Data Index: 1364260 (Python Code)

Max Activating Token Index: 6

Text #12

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.2997. Min Act: -0.0001

Data Index: 1653137 (Python Code)

Max Activating Token Index: 627

Text #13

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.2714. Min Act: -0.0001

Data Index: 1584925 (Python Code)

Max Activating Token Index: 287

Text #14

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.3195. Min Act: -0.0001

Data Index: 591742 (C4 (Web Text))

Max Activating Token Index: 22

Text #15

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.2842. Min Act: -0.0001

Data Index: 69124 (C4 (Web Text))

Max Activating Token Index: 983

Text #16

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.2665. Min Act: -0.0001

Data Index: 1530652 (Python Code)

Max Activating Token Index: 34

Text #17

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.2805. Min Act: -0.0001

Data Index: 1083612 (C4 (Web Text))

Max Activating Token Index: 784

Text #18

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.2676. Min Act: -0.0001

Data Index: 1652235 (Python Code)

Max Activating Token Index: 434

Text #19

Max Range: 0.3895. Min Range: -0.3895

Max Act: 0.3139. Min Act: -0.0001

Data Index: 1468864 (Python Code)

Max Activating Token Index: 606