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

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

Neuron 775 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.7115. Min Act: -0.0000

Data Index: 1435108 (Python Code)

Max Activating Token Index: 649

Text #1

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.4677. Min Act: -0.0000

Data Index: 1496598 (Python Code)

Max Activating Token Index: 773

Text #2

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.4351. Min Act: -0.0000

Data Index: 444826 (C4 (Web Text))

Max Activating Token Index: 456

Text #3

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.4232. Min Act: -0.0000

Data Index: 191084 (C4 (Web Text))

Max Activating Token Index: 109

Text #4

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.4311. Min Act: -0.0000

Data Index: 1459446 (Python Code)

Max Activating Token Index: 898

Text #5

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.3928. Min Act: -0.0000

Data Index: 654850 (C4 (Web Text))

Max Activating Token Index: 854

Text #6

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.3587. Min Act: -0.0000

Data Index: 1557882 (Python Code)

Max Activating Token Index: 520

Text #7

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.3627. Min Act: -0.0000

Data Index: 635413 (C4 (Web Text))

Max Activating Token Index: 615

Text #8

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.3618. Min Act: -0.0000

Data Index: 666214 (C4 (Web Text))

Max Activating Token Index: 33

Text #9

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.3450. Min Act: -0.0000

Data Index: 1503298 (Python Code)

Max Activating Token Index: 229

Text #10

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.3226. Min Act: -0.0000

Data Index: 1468915 (Python Code)

Max Activating Token Index: 992

Text #11

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.3186. Min Act: -0.0000

Data Index: 952611 (C4 (Web Text))

Max Activating Token Index: 142

Text #12

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.3109. Min Act: -0.0000

Data Index: 1331343 (C4 (Web Text))

Max Activating Token Index: 883

Text #13

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.3208. Min Act: -0.0000

Data Index: 1390142 (Python Code)

Max Activating Token Index: 162

Text #14

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.3012. Min Act: -0.0000

Data Index: 70958 (C4 (Web Text))

Max Activating Token Index: 612

Text #15

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.3073. Min Act: -0.0000

Data Index: 988518 (C4 (Web Text))

Max Activating Token Index: 106

Text #16

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.3080. Min Act: -0.0000

Data Index: 1370375 (Python Code)

Max Activating Token Index: 804

Text #17

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.2815. Min Act: -0.0000

Data Index: 103004 (C4 (Web Text))

Max Activating Token Index: 104

Text #18

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.2807. Min Act: -0.0000

Data Index: 814722 (C4 (Web Text))

Max Activating Token Index: 681

Text #19

Max Range: 0.7115. Min Range: -0.7115

Max Act: 0.2864. Min Act: -0.0000

Data Index: 517658 (C4 (Web Text))

Max Activating Token Index: 279