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

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

Neuron 1663 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.3636. Min Range: -3.3636

Max Act: 3.1852. Min Act: -0.0001

Data Index: 17612 (C4 (Web Text))

Max Activating Token Index: 517

Text #1

Max Range: 3.3636. Min Range: -3.3636

Max Act: 3.3636. Min Act: -0.0001

Data Index: 783661 (C4 (Web Text))

Max Activating Token Index: 91

Text #2

Max Range: 3.3636. Min Range: -3.3636

Max Act: 3.2425. Min Act: -0.0001

Data Index: 1126174 (C4 (Web Text))

Max Activating Token Index: 459

Text #3

Max Range: 3.3636. Min Range: -3.3636

Max Act: 3.0744. Min Act: -0.0001

Data Index: 104708 (C4 (Web Text))

Max Activating Token Index: 491

Text #4

Max Range: 3.3636. Min Range: -3.3636

Max Act: 3.0334. Min Act: -0.0001

Data Index: 1284838 (C4 (Web Text))

Max Activating Token Index: 533

Text #5

Max Range: 3.3636. Min Range: -3.3636

Max Act: 3.2112. Min Act: -0.0001

Data Index: 141212 (C4 (Web Text))

Max Activating Token Index: 195

Text #6

Max Range: 3.3636. Min Range: -3.3636

Max Act: 3.1541. Min Act: -0.0001

Data Index: 669718 (C4 (Web Text))

Max Activating Token Index: 114

Text #7

Max Range: 3.3636. Min Range: -3.3636

Max Act: 3.3209. Min Act: -0.0001

Data Index: 523625 (C4 (Web Text))

Max Activating Token Index: 347

Text #8

Max Range: 3.3636. Min Range: -3.3636

Max Act: 3.0690. Min Act: -0.0001

Data Index: 250228 (C4 (Web Text))

Max Activating Token Index: 21

Text #9

Max Range: 3.3636. Min Range: -3.3636

Max Act: 3.0774. Min Act: -0.0001

Data Index: 262011 (C4 (Web Text))

Max Activating Token Index: 156

Text #10

Max Range: 3.3636. Min Range: -3.3636

Max Act: 3.0046. Min Act: -0.0001

Data Index: 198340 (C4 (Web Text))

Max Activating Token Index: 34

Text #11

Max Range: 3.3636. Min Range: -3.3636

Max Act: 2.9986. Min Act: -0.0001

Data Index: 822694 (C4 (Web Text))

Max Activating Token Index: 308

Text #12

Max Range: 3.3636. Min Range: -3.3636

Max Act: 2.9954. Min Act: -0.0001

Data Index: 936450 (C4 (Web Text))

Max Activating Token Index: 36

Text #13

Max Range: 3.3636. Min Range: -3.3636

Max Act: 3.0861. Min Act: -0.0001

Data Index: 511214 (C4 (Web Text))

Max Activating Token Index: 107

Text #14

Max Range: 3.3636. Min Range: -3.3636

Max Act: 2.6947. Min Act: -0.0001

Data Index: 129761 (C4 (Web Text))

Max Activating Token Index: 706

Text #15

Max Range: 3.3636. Min Range: -3.3636

Max Act: 2.9696. Min Act: -0.0001

Data Index: 776726 (C4 (Web Text))

Max Activating Token Index: 8

Text #16

Max Range: 3.3636. Min Range: -3.3636

Max Act: 2.9132. Min Act: -0.0001

Data Index: 868360 (C4 (Web Text))

Max Activating Token Index: 528

Text #17

Max Range: 3.3636. Min Range: -3.3636

Max Act: 2.7906. Min Act: -0.0001

Data Index: 398881 (C4 (Web Text))

Max Activating Token Index: 799

Text #18

Max Range: 3.3636. Min Range: -3.3636

Max Act: 2.9164. Min Act: -0.0001

Data Index: 1347669 (C4 (Web Text))

Max Activating Token Index: 293

Text #19

Max Range: 3.3636. Min Range: -3.3636

Max Act: 2.7568. Min Act: -0.0001

Data Index: 253357 (C4 (Web Text))

Max Activating Token Index: 83