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

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

Neuron 4442 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.7761. Min Range: -3.7761

Max Act: 3.7761. Min Act: -0.0001

Data Index: 709930 (C4 (Web Text))

Max Activating Token Index: 612

Text #1

Max Range: 3.7761. Min Range: -3.7761

Max Act: 3.3979. Min Act: -0.0001

Data Index: 798474 (C4 (Web Text))

Max Activating Token Index: 38

Text #2

Max Range: 3.7761. Min Range: -3.7761

Max Act: 3.3316. Min Act: -0.0001

Data Index: 173817 (C4 (Web Text))

Max Activating Token Index: 25

Text #3

Max Range: 3.7761. Min Range: -3.7761

Max Act: 3.1895. Min Act: -0.0001

Data Index: 118464 (C4 (Web Text))

Max Activating Token Index: 78

Text #4

Max Range: 3.7761. Min Range: -3.7761

Max Act: 3.3052. Min Act: -0.0001

Data Index: 23457 (C4 (Web Text))

Max Activating Token Index: 612

Text #5

Max Range: 3.7761. Min Range: -3.7761

Max Act: 3.3501. Min Act: -0.0001

Data Index: 1240506 (C4 (Web Text))

Max Activating Token Index: 81

Text #6

Max Range: 3.7761. Min Range: -3.7761

Max Act: 3.1817. Min Act: -0.0001

Data Index: 1526040 (Python Code)

Max Activating Token Index: 276

Text #7

Max Range: 3.7761. Min Range: -3.7761

Max Act: 3.2075. Min Act: -0.0001

Data Index: 490984 (C4 (Web Text))

Max Activating Token Index: 936

Text #8

Max Range: 3.7761. Min Range: -3.7761

Max Act: 2.7968. Min Act: -0.0001

Data Index: 1480071 (Python Code)

Max Activating Token Index: 373

Text #9

Max Range: 3.7761. Min Range: -3.7761

Max Act: 2.8870. Min Act: -0.0001

Data Index: 1106793 (C4 (Web Text))

Max Activating Token Index: 844

Text #10

Max Range: 3.7761. Min Range: -3.7761

Max Act: 2.9835. Min Act: -0.0001

Data Index: 1141838 (C4 (Web Text))

Max Activating Token Index: 531

Text #11

Max Range: 3.7761. Min Range: -3.7761

Max Act: 3.0893. Min Act: -0.0001

Data Index: 298821 (C4 (Web Text))

Max Activating Token Index: 31

Text #12

Max Range: 3.7761. Min Range: -3.7761

Max Act: 3.0829. Min Act: -0.0001

Data Index: 1155117 (C4 (Web Text))

Max Activating Token Index: 399

Text #13

Max Range: 3.7761. Min Range: -3.7761

Max Act: 2.7663. Min Act: -0.0001

Data Index: 1449710 (Python Code)

Max Activating Token Index: 174

Text #14

Max Range: 3.7761. Min Range: -3.7761

Max Act: 3.1415. Min Act: -0.0001

Data Index: 373735 (C4 (Web Text))

Max Activating Token Index: 158

Text #15

Max Range: 3.7761. Min Range: -3.7761

Max Act: 2.9988. Min Act: -0.0001

Data Index: 1303733 (C4 (Web Text))

Max Activating Token Index: 969

Text #16

Max Range: 3.7761. Min Range: -3.7761

Max Act: 2.5444. Min Act: -0.0001

Data Index: 574006 (C4 (Web Text))

Max Activating Token Index: 750

Text #17

Max Range: 3.7761. Min Range: -3.7761

Max Act: 2.6541. Min Act: -0.0001

Data Index: 1229117 (C4 (Web Text))

Max Activating Token Index: 854

Text #18

Max Range: 3.7761. Min Range: -3.7761

Max Act: 2.7065. Min Act: -0.0001

Data Index: 344184 (C4 (Web Text))

Max Activating Token Index: 59

Text #19

Max Range: 3.7761. Min Range: -3.7761

Max Act: 2.7902. Min Act: -0.0001

Data Index: 522912 (C4 (Web Text))

Max Activating Token Index: 927