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

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

Neuron 3650 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.9252. Min Act: -0.0003

Data Index: 756265 (C4 (Web Text))

Max Activating Token Index: 467

Text #1

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.9105. Min Act: -0.0003

Data Index: 890345 (C4 (Web Text))

Max Activating Token Index: 689

Text #2

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.8208. Min Act: -0.0003

Data Index: 532781 (C4 (Web Text))

Max Activating Token Index: 282

Text #3

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.8182. Min Act: -0.0003

Data Index: 874235 (C4 (Web Text))

Max Activating Token Index: 79

Text #4

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.8381. Min Act: -0.0003

Data Index: 234883 (C4 (Web Text))

Max Activating Token Index: 177

Text #5

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.8383. Min Act: -0.0003

Data Index: 1260481 (C4 (Web Text))

Max Activating Token Index: 328

Text #6

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.7995. Min Act: -0.0003

Data Index: 116490 (C4 (Web Text))

Max Activating Token Index: 232

Text #7

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.7964. Min Act: -0.0003

Data Index: 1091706 (C4 (Web Text))

Max Activating Token Index: 160

Text #8

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.7961. Min Act: -0.0003

Data Index: 1242374 (C4 (Web Text))

Max Activating Token Index: 297

Text #9

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.7448. Min Act: -0.0002

Data Index: 893707 (C4 (Web Text))

Max Activating Token Index: 717

Text #10

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.7656. Min Act: -0.0003

Data Index: 957098 (C4 (Web Text))

Max Activating Token Index: 750

Text #11

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.7134. Min Act: -0.0003

Data Index: 1122600 (C4 (Web Text))

Max Activating Token Index: 289

Text #12

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.7300. Min Act: -0.0003

Data Index: 878512 (C4 (Web Text))

Max Activating Token Index: 648

Text #13

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.7143. Min Act: -0.0003

Data Index: 310947 (C4 (Web Text))

Max Activating Token Index: 620

Text #14

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.6665. Min Act: -0.0003

Data Index: 41909 (C4 (Web Text))

Max Activating Token Index: 630

Text #15

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.6895. Min Act: -0.0003

Data Index: 972272 (C4 (Web Text))

Max Activating Token Index: 712

Text #16

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.6836. Min Act: -0.0003

Data Index: 911478 (C4 (Web Text))

Max Activating Token Index: 579

Text #17

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.6779. Min Act: -0.0003

Data Index: 632981 (C4 (Web Text))

Max Activating Token Index: 738

Text #18

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.6536. Min Act: -0.0002

Data Index: 979077 (C4 (Web Text))

Max Activating Token Index: 864

Text #19

Max Range: 0.9252. Min Range: -0.9252

Max Act: 0.6725. Min Act: -0.0003

Data Index: 875763 (C4 (Web Text))

Max Activating Token Index: 99