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

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

Neuron 1239 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: 2.7270. Min Range: -2.7270

Max Act: 2.6897. Min Act: -0.0000

Data Index: 1649726 (Python Code)

Max Activating Token Index: 19

Text #1

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.6429. Min Act: -0.0000

Data Index: 964221 (C4 (Web Text))

Max Activating Token Index: 851

Text #2

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.7270. Min Act: -0.0000

Data Index: 960199 (C4 (Web Text))

Max Activating Token Index: 642

Text #3

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.5341. Min Act: -0.0000

Data Index: 827878 (C4 (Web Text))

Max Activating Token Index: 644

Text #4

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.3084. Min Act: -0.0000

Data Index: 1395354 (Python Code)

Max Activating Token Index: 25

Text #5

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.2926. Min Act: -0.0000

Data Index: 772205 (C4 (Web Text))

Max Activating Token Index: 470

Text #6

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.3324. Min Act: -0.0000

Data Index: 725759 (C4 (Web Text))

Max Activating Token Index: 505

Text #7

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.2035. Min Act: -0.0000

Data Index: 251015 (C4 (Web Text))

Max Activating Token Index: 345

Text #8

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.2493. Min Act: -0.0000

Data Index: 1633778 (Python Code)

Max Activating Token Index: 171

Text #9

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.2328. Min Act: -0.0000

Data Index: 1569401 (Python Code)

Max Activating Token Index: 209

Text #10

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.1139. Min Act: -0.0000

Data Index: 810226 (C4 (Web Text))

Max Activating Token Index: 629

Text #11

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.1020. Min Act: -0.0000

Data Index: 291734 (C4 (Web Text))

Max Activating Token Index: 983

Text #12

Max Range: 2.7270. Min Range: -2.7270

Max Act: 1.9075. Min Act: -0.0000

Data Index: 1393553 (Python Code)

Max Activating Token Index: 144

Text #13

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.0376. Min Act: -0.0000

Data Index: 850500 (C4 (Web Text))

Max Activating Token Index: 1014

Text #14

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.1539. Min Act: -0.0000

Data Index: 855306 (C4 (Web Text))

Max Activating Token Index: 228

Text #15

Max Range: 2.7270. Min Range: -2.7270

Max Act: 2.0307. Min Act: -0.0000

Data Index: 1525911 (Python Code)

Max Activating Token Index: 472

Text #16

Max Range: 2.7270. Min Range: -2.7270

Max Act: 1.9851. Min Act: -0.0000

Data Index: 569848 (C4 (Web Text))

Max Activating Token Index: 596

Text #17

Max Range: 2.7270. Min Range: -2.7270

Max Act: 1.8935. Min Act: -0.0000

Data Index: 1037095 (C4 (Web Text))

Max Activating Token Index: 515

Text #18

Max Range: 2.7270. Min Range: -2.7270

Max Act: 1.8049. Min Act: -0.0000

Data Index: 57329 (C4 (Web Text))

Max Activating Token Index: 287

Text #19

Max Range: 2.7270. Min Range: -2.7270

Max Act: 1.8442. Min Act: -0.0000

Data Index: 234043 (C4 (Web Text))

Max Activating Token Index: 548