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

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

Neuron 2015 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0809. Min Act: -0.0000

Data Index: 441470 (C4 (Web Text))

Max Activating Token Index: 344

Text #1

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0644. Min Act: -0.0000

Data Index: 1536199 (Python Code)

Max Activating Token Index: 221

Text #2

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0708. Min Act: -0.0001

Data Index: 121956 (C4 (Web Text))

Max Activating Token Index: 32

Text #3

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0650. Min Act: -0.0000

Data Index: 1600534 (Python Code)

Max Activating Token Index: 254

Text #4

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0651. Min Act: -0.0001

Data Index: 780105 (C4 (Web Text))

Max Activating Token Index: 526

Text #5

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0688. Min Act: -0.0000

Data Index: 1172605 (C4 (Web Text))

Max Activating Token Index: 551

Text #6

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0637. Min Act: -0.0001

Data Index: 453023 (C4 (Web Text))

Max Activating Token Index: 186

Text #7

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0676. Min Act: -0.0000

Data Index: 1556572 (Python Code)

Max Activating Token Index: 396

Text #8

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0621. Min Act: -0.0000

Data Index: 1448440 (Python Code)

Max Activating Token Index: 255

Text #9

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0671. Min Act: -0.0000

Data Index: 1277795 (C4 (Web Text))

Max Activating Token Index: 738

Text #10

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0616. Min Act: -0.0000

Data Index: 111422 (C4 (Web Text))

Max Activating Token Index: 318

Text #11

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0652. Min Act: -0.0000

Data Index: 1497748 (Python Code)

Max Activating Token Index: 310

Text #12

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0643. Min Act: -0.0000

Data Index: 1385118 (Python Code)

Max Activating Token Index: 923

Text #13

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0574. Min Act: -0.0000

Data Index: 1277234 (C4 (Web Text))

Max Activating Token Index: 696

Text #14

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0594. Min Act: -0.0000

Data Index: 1524950 (Python Code)

Max Activating Token Index: 174

Text #15

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0567. Min Act: -0.0001

Data Index: 319679 (C4 (Web Text))

Max Activating Token Index: 362

Text #16

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0622. Min Act: -0.0000

Data Index: 22149 (C4 (Web Text))

Max Activating Token Index: 754

Text #17

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0597. Min Act: -0.0000

Data Index: 16537 (C4 (Web Text))

Max Activating Token Index: 35

Text #18

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0591. Min Act: -0.0000

Data Index: 1371458 (Python Code)

Max Activating Token Index: 177

Text #19

Max Range: 0.0809. Min Range: -0.0809

Max Act: 0.0591. Min Act: -0.0000

Data Index: 1570053 (Python Code)

Max Activating Token Index: 64