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

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

Neuron 808 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.1785. Min Range: -1.1785

Max Act: 1.1785. Min Act: -0.0001

Data Index: 408500 (C4 (Web Text))

Max Activating Token Index: 369

Text #1

Max Range: 1.1785. Min Range: -1.1785

Max Act: 1.0962. Min Act: -0.0001

Data Index: 972203 (C4 (Web Text))

Max Activating Token Index: 150

Text #2

Max Range: 1.1785. Min Range: -1.1785

Max Act: 1.0388. Min Act: -0.0001

Data Index: 677930 (C4 (Web Text))

Max Activating Token Index: 923

Text #3

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.9761. Min Act: -0.0001

Data Index: 1104468 (C4 (Web Text))

Max Activating Token Index: 98

Text #4

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.9447. Min Act: -0.0001

Data Index: 1358350 (C4 (Web Text))

Max Activating Token Index: 748

Text #5

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.9531. Min Act: -0.0001

Data Index: 616186 (C4 (Web Text))

Max Activating Token Index: 64

Text #6

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.9207. Min Act: -0.0001

Data Index: 830058 (C4 (Web Text))

Max Activating Token Index: 485

Text #7

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.8371. Min Act: -0.0001

Data Index: 1321616 (C4 (Web Text))

Max Activating Token Index: 64

Text #8

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.8156. Min Act: -0.0001

Data Index: 1025998 (C4 (Web Text))

Max Activating Token Index: 149

Text #9

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.9018. Min Act: -0.0001

Data Index: 1117553 (C4 (Web Text))

Max Activating Token Index: 598

Text #10

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.8851. Min Act: -0.0001

Data Index: 269146 (C4 (Web Text))

Max Activating Token Index: 99

Text #11

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.8251. Min Act: -0.0001

Data Index: 111622 (C4 (Web Text))

Max Activating Token Index: 768

Text #12

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.7747. Min Act: -0.0001

Data Index: 303134 (C4 (Web Text))

Max Activating Token Index: 704

Text #13

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.8025. Min Act: -0.0001

Data Index: 757331 (C4 (Web Text))

Max Activating Token Index: 351

Text #14

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.8018. Min Act: -0.0001

Data Index: 1635027 (Python Code)

Max Activating Token Index: 221

Text #15

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.8255. Min Act: -0.0001

Data Index: 999337 (C4 (Web Text))

Max Activating Token Index: 397

Text #16

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.8173. Min Act: -0.0001

Data Index: 1088317 (C4 (Web Text))

Max Activating Token Index: 289

Text #17

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.8075. Min Act: -0.0001

Data Index: 416543 (C4 (Web Text))

Max Activating Token Index: 886

Text #18

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.7652. Min Act: -0.0001

Data Index: 258373 (C4 (Web Text))

Max Activating Token Index: 776

Text #19

Max Range: 1.1785. Min Range: -1.1785

Max Act: 0.7198. Min Act: -0.0001

Data Index: 485959 (C4 (Web Text))

Max Activating Token Index: 196