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

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

Neuron 923 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.4340. Min Act: -0.0000

Data Index: 1234377 (C4 (Web Text))

Max Activating Token Index: 776

Text #1

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.4157. Min Act: -0.0000

Data Index: 786098 (C4 (Web Text))

Max Activating Token Index: 860

Text #2

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.4383. Min Act: -0.0000

Data Index: 1648838 (Python Code)

Max Activating Token Index: 934

Text #3

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.4026. Min Act: -0.0000

Data Index: 1506606 (Python Code)

Max Activating Token Index: 290

Text #4

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3856. Min Act: -0.0000

Data Index: 1430417 (Python Code)

Max Activating Token Index: 806

Text #5

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3688. Min Act: -0.0000

Data Index: 107706 (C4 (Web Text))

Max Activating Token Index: 876

Text #6

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3794. Min Act: -0.0000

Data Index: 414881 (C4 (Web Text))

Max Activating Token Index: 207

Text #7

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3817. Min Act: -0.0000

Data Index: 1611119 (Python Code)

Max Activating Token Index: 78

Text #8

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3868. Min Act: -0.0000

Data Index: 486856 (C4 (Web Text))

Max Activating Token Index: 449

Text #9

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3631. Min Act: -0.0000

Data Index: 1277042 (C4 (Web Text))

Max Activating Token Index: 386

Text #10

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3582. Min Act: -0.0000

Data Index: 1090239 (C4 (Web Text))

Max Activating Token Index: 443

Text #11

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3529. Min Act: -0.0000

Data Index: 1599885 (Python Code)

Max Activating Token Index: 456

Text #12

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3564. Min Act: -0.0000

Data Index: 373143 (C4 (Web Text))

Max Activating Token Index: 166

Text #13

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3300. Min Act: -0.0000

Data Index: 323452 (C4 (Web Text))

Max Activating Token Index: 581

Text #14

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3668. Min Act: -0.0000

Data Index: 1619777 (Python Code)

Max Activating Token Index: 116

Text #15

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3405. Min Act: -0.0000

Data Index: 1390840 (Python Code)

Max Activating Token Index: 297

Text #16

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3507. Min Act: -0.0000

Data Index: 537848 (C4 (Web Text))

Max Activating Token Index: 889

Text #17

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3387. Min Act: -0.0000

Data Index: 1302936 (C4 (Web Text))

Max Activating Token Index: 608

Text #18

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3124. Min Act: -0.0000

Data Index: 1558919 (Python Code)

Max Activating Token Index: 1023

Text #19

Max Range: 0.4383. Min Range: -0.4383

Max Act: 0.3428. Min Act: -0.0000

Data Index: 390847 (C4 (Web Text))

Max Activating Token Index: 862