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

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

Neuron 591 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: 1.5254. Min Range: -1.5254

Max Act: 1.5254. Min Act: -0.0000

Data Index: 402493 (C4 (Web Text))

Max Activating Token Index: 232

Text #1

Max Range: 1.5254. Min Range: -1.5254

Max Act: 1.1123. Min Act: -0.0000

Data Index: 1245668 (C4 (Web Text))

Max Activating Token Index: 413

Text #2

Max Range: 1.5254. Min Range: -1.5254

Max Act: 1.0859. Min Act: -0.0000

Data Index: 1228603 (C4 (Web Text))

Max Activating Token Index: 324

Text #3

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.9780. Min Act: -0.0000

Data Index: 332172 (C4 (Web Text))

Max Activating Token Index: 126

Text #4

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.9747. Min Act: -0.0000

Data Index: 1138822 (C4 (Web Text))

Max Activating Token Index: 978

Text #5

Max Range: 1.5254. Min Range: -1.5254

Max Act: 1.0036. Min Act: -0.0000

Data Index: 1136902 (C4 (Web Text))

Max Activating Token Index: 124

Text #6

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.8915. Min Act: -0.0000

Data Index: 441799 (C4 (Web Text))

Max Activating Token Index: 479

Text #7

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.8766. Min Act: -0.0000

Data Index: 1637368 (Python Code)

Max Activating Token Index: 70

Text #8

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.8545. Min Act: -0.0000

Data Index: 905392 (C4 (Web Text))

Max Activating Token Index: 646

Text #9

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.9030. Min Act: -0.0000

Data Index: 917060 (C4 (Web Text))

Max Activating Token Index: 124

Text #10

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.8349. Min Act: -0.0000

Data Index: 360847 (C4 (Web Text))

Max Activating Token Index: 19

Text #11

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.8428. Min Act: -0.0000

Data Index: 1399393 (Python Code)

Max Activating Token Index: 153

Text #12

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.8133. Min Act: -0.0000

Data Index: 66386 (C4 (Web Text))

Max Activating Token Index: 38

Text #13

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.7653. Min Act: -0.0000

Data Index: 280740 (C4 (Web Text))

Max Activating Token Index: 470

Text #14

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.8185. Min Act: -0.0000

Data Index: 1394127 (Python Code)

Max Activating Token Index: 524

Text #15

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.8052. Min Act: -0.0000

Data Index: 1419105 (Python Code)

Max Activating Token Index: 349

Text #16

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.7984. Min Act: -0.0000

Data Index: 1450799 (Python Code)

Max Activating Token Index: 426

Text #17

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.7829. Min Act: -0.0000

Data Index: 1412246 (Python Code)

Max Activating Token Index: 120

Text #18

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.8385. Min Act: -0.0000

Data Index: 330097 (C4 (Web Text))

Max Activating Token Index: 677

Text #19

Max Range: 1.5254. Min Range: -1.5254

Max Act: 0.7384. Min Act: -0.0000

Data Index: 1425566 (Python Code)

Max Activating Token Index: 300