Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 2364 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 5.7660. Min Range: -5.7660

Max Act: 5.7660. Min Act: -0.0001

Data Index: 691373 (C4 (Web Text))

Max Activating Token Index: 930

Text #1

Max Range: 5.7660. Min Range: -5.7660

Max Act: 5.2643. Min Act: -0.0001

Data Index: 587867 (C4 (Web Text))

Max Activating Token Index: 49

Text #2

Max Range: 5.7660. Min Range: -5.7660

Max Act: 5.1588. Min Act: -0.0001

Data Index: 7894 (C4 (Web Text))

Max Activating Token Index: 525

Text #3

Max Range: 5.7660. Min Range: -5.7660

Max Act: 4.7371. Min Act: -0.0001

Data Index: 404578 (C4 (Web Text))

Max Activating Token Index: 573

Text #4

Max Range: 5.7660. Min Range: -5.7660

Max Act: 4.6428. Min Act: -0.0001

Data Index: 224160 (C4 (Web Text))

Max Activating Token Index: 731

Text #5

Max Range: 5.7660. Min Range: -5.7660

Max Act: 4.5406. Min Act: -0.0001

Data Index: 512546 (C4 (Web Text))

Max Activating Token Index: 158

Text #6

Max Range: 5.7660. Min Range: -5.7660

Max Act: 4.5224. Min Act: -0.0001

Data Index: 912329 (C4 (Web Text))

Max Activating Token Index: 860

Text #7

Max Range: 5.7660. Min Range: -5.7660

Max Act: 4.3550. Min Act: -0.0001

Data Index: 62573 (C4 (Web Text))

Max Activating Token Index: 538

Text #8

Max Range: 5.7660. Min Range: -5.7660

Max Act: 4.3184. Min Act: -0.0001

Data Index: 1263145 (C4 (Web Text))

Max Activating Token Index: 996

Text #9

Max Range: 5.7660. Min Range: -5.7660

Max Act: 4.0156. Min Act: -0.0001

Data Index: 1165017 (C4 (Web Text))

Max Activating Token Index: 474

Text #10

Max Range: 5.7660. Min Range: -5.7660

Max Act: 3.9682. Min Act: -0.0001

Data Index: 58649 (C4 (Web Text))

Max Activating Token Index: 473

Text #11

Max Range: 5.7660. Min Range: -5.7660

Max Act: 4.0623. Min Act: -0.0001

Data Index: 120713 (C4 (Web Text))

Max Activating Token Index: 577

Text #12

Max Range: 5.7660. Min Range: -5.7660

Max Act: 3.8236. Min Act: -0.0001

Data Index: 206262 (C4 (Web Text))

Max Activating Token Index: 511

Text #13

Max Range: 5.7660. Min Range: -5.7660

Max Act: 3.9912. Min Act: -0.0001

Data Index: 769226 (C4 (Web Text))

Max Activating Token Index: 600

Text #14

Max Range: 5.7660. Min Range: -5.7660

Max Act: 3.7843. Min Act: -0.0001

Data Index: 65471 (C4 (Web Text))

Max Activating Token Index: 748

Text #15

Max Range: 5.7660. Min Range: -5.7660

Max Act: 3.7012. Min Act: -0.0001

Data Index: 585345 (C4 (Web Text))

Max Activating Token Index: 636

Text #16

Max Range: 5.7660. Min Range: -5.7660

Max Act: 3.4627. Min Act: -0.0001

Data Index: 1302437 (C4 (Web Text))

Max Activating Token Index: 643

Text #17

Max Range: 5.7660. Min Range: -5.7660

Max Act: 3.7303. Min Act: -0.0001

Data Index: 281622 (C4 (Web Text))

Max Activating Token Index: 340

Text #18

Max Range: 5.7660. Min Range: -5.7660

Max Act: 3.6827. Min Act: -0.0001

Data Index: 1107845 (C4 (Web Text))

Max Activating Token Index: 518

Text #19

Max Range: 5.7660. Min Range: -5.7660

Max Act: 3.5081. Min Act: -0.0001

Data Index: 769215 (C4 (Web Text))

Max Activating Token Index: 756