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

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

Neuron 1092 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.5164. Min Act: -0.0001

Data Index: 461114 (C4 (Web Text))

Max Activating Token Index: 259

Text #1

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4915. Min Act: -0.0000

Data Index: 482545 (C4 (Web Text))

Max Activating Token Index: 606

Text #2

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4420. Min Act: -0.0001

Data Index: 382853 (C4 (Web Text))

Max Activating Token Index: 5

Text #3

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4556. Min Act: -0.0001

Data Index: 313283 (C4 (Web Text))

Max Activating Token Index: 203

Text #4

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4390. Min Act: -0.0000

Data Index: 1226283 (C4 (Web Text))

Max Activating Token Index: 46

Text #5

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4684. Min Act: -0.0001

Data Index: 811915 (C4 (Web Text))

Max Activating Token Index: 369

Text #6

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4390. Min Act: -0.0001

Data Index: 1466854 (Python Code)

Max Activating Token Index: 946

Text #7

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4625. Min Act: -0.0001

Data Index: 295956 (C4 (Web Text))

Max Activating Token Index: 296

Text #8

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4320. Min Act: -0.0000

Data Index: 1142340 (C4 (Web Text))

Max Activating Token Index: 231

Text #9

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4597. Min Act: -0.0000

Data Index: 179540 (C4 (Web Text))

Max Activating Token Index: 383

Text #10

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4418. Min Act: -0.0000

Data Index: 197726 (C4 (Web Text))

Max Activating Token Index: 111

Text #11

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4543. Min Act: -0.0001

Data Index: 1392438 (Python Code)

Max Activating Token Index: 193

Text #12

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4546. Min Act: -0.0001

Data Index: 776193 (C4 (Web Text))

Max Activating Token Index: 29

Text #13

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4399. Min Act: -0.0000

Data Index: 813269 (C4 (Web Text))

Max Activating Token Index: 460

Text #14

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4388. Min Act: -0.0001

Data Index: 1459137 (Python Code)

Max Activating Token Index: 455

Text #15

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4193. Min Act: -0.0000

Data Index: 417459 (C4 (Web Text))

Max Activating Token Index: 27

Text #16

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4123. Min Act: -0.0000

Data Index: 1291565 (C4 (Web Text))

Max Activating Token Index: 235

Text #17

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4243. Min Act: -0.0000

Data Index: 1232951 (C4 (Web Text))

Max Activating Token Index: 760

Text #18

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4191. Min Act: -0.0001

Data Index: 1369457 (Python Code)

Max Activating Token Index: 168

Text #19

Max Range: 0.5164. Min Range: -0.5164

Max Act: 0.4042. Min Act: -0.0000

Data Index: 905322 (C4 (Web Text))

Max Activating Token Index: 870