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

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

Neuron 326 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 3.1248. Min Range: -3.1248

Max Act: 3.1248. Min Act: -0.0000

Data Index: 552150 (C4 (Web Text))

Max Activating Token Index: 174

Text #1

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.5786. Min Act: -0.0000

Data Index: 797944 (C4 (Web Text))

Max Activating Token Index: 947

Text #2

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.6690. Min Act: -0.0000

Data Index: 90049 (C4 (Web Text))

Max Activating Token Index: 665

Text #3

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.6190. Min Act: -0.0000

Data Index: 524962 (C4 (Web Text))

Max Activating Token Index: 697

Text #4

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.4137. Min Act: -0.0000

Data Index: 969606 (C4 (Web Text))

Max Activating Token Index: 481

Text #5

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.2938. Min Act: -0.0000

Data Index: 2800 (C4 (Web Text))

Max Activating Token Index: 1018

Text #6

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.2957. Min Act: -0.0000

Data Index: 1634506 (Python Code)

Max Activating Token Index: 496

Text #7

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.2901. Min Act: -0.0000

Data Index: 815774 (C4 (Web Text))

Max Activating Token Index: 473

Text #8

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.2172. Min Act: -0.0000

Data Index: 799845 (C4 (Web Text))

Max Activating Token Index: 322

Text #9

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.2121. Min Act: -0.0000

Data Index: 590001 (C4 (Web Text))

Max Activating Token Index: 821

Text #10

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.0840. Min Act: -0.0000

Data Index: 1603585 (Python Code)

Max Activating Token Index: 578

Text #11

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.1622. Min Act: -0.0000

Data Index: 274461 (C4 (Web Text))

Max Activating Token Index: 352

Text #12

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.1747. Min Act: -0.0000

Data Index: 1438861 (Python Code)

Max Activating Token Index: 160

Text #13

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.1551. Min Act: -0.0000

Data Index: 763746 (C4 (Web Text))

Max Activating Token Index: 459

Text #14

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.1138. Min Act: -0.0000

Data Index: 1512839 (Python Code)

Max Activating Token Index: 492

Text #15

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.0917. Min Act: -0.0000

Data Index: 1447562 (Python Code)

Max Activating Token Index: 68

Text #16

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.0934. Min Act: -0.0000

Data Index: 1500033 (Python Code)

Max Activating Token Index: 1008

Text #17

Max Range: 3.1248. Min Range: -3.1248

Max Act: 1.9623. Min Act: -0.0000

Data Index: 1286900 (C4 (Web Text))

Max Activating Token Index: 220

Text #18

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.0631. Min Act: -0.0000

Data Index: 345274 (C4 (Web Text))

Max Activating Token Index: 349

Text #19

Max Range: 3.1248. Min Range: -3.1248

Max Act: 2.0431. Min Act: -0.0000

Data Index: 1431551 (Python Code)

Max Activating Token Index: 979