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

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

Neuron 326 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: 0.3022. Min Range: -0.3022

Max Act: 0.3022. Min Act: -0.0001

Data Index: 608952 (C4 (Web Text))

Max Activating Token Index: 933

Text #1

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.3001. Min Act: -0.0001

Data Index: 1087026 (C4 (Web Text))

Max Activating Token Index: 171

Text #2

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2672. Min Act: -0.0001

Data Index: 772762 (C4 (Web Text))

Max Activating Token Index: 679

Text #3

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2856. Min Act: -0.0001

Data Index: 307532 (C4 (Web Text))

Max Activating Token Index: 1023

Text #4

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2669. Min Act: -0.0001

Data Index: 594387 (C4 (Web Text))

Max Activating Token Index: 873

Text #5

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2481. Min Act: -0.0001

Data Index: 33369 (C4 (Web Text))

Max Activating Token Index: 73

Text #6

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2545. Min Act: -0.0001

Data Index: 514264 (C4 (Web Text))

Max Activating Token Index: 903

Text #7

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2279. Min Act: -0.0001

Data Index: 1525176 (Python Code)

Max Activating Token Index: 834

Text #8

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2109. Min Act: -0.0001

Data Index: 426436 (C4 (Web Text))

Max Activating Token Index: 91

Text #9

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2414. Min Act: -0.0001

Data Index: 1368935 (Python Code)

Max Activating Token Index: 100

Text #10

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2668. Min Act: -0.0001

Data Index: 1074077 (C4 (Web Text))

Max Activating Token Index: 962

Text #11

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2174. Min Act: -0.0001

Data Index: 555828 (C4 (Web Text))

Max Activating Token Index: 451

Text #12

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2192. Min Act: -0.0001

Data Index: 1587405 (Python Code)

Max Activating Token Index: 330

Text #13

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2233. Min Act: -0.0001

Data Index: 268623 (C4 (Web Text))

Max Activating Token Index: 866

Text #14

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2350. Min Act: -0.0001

Data Index: 697189 (C4 (Web Text))

Max Activating Token Index: 398

Text #15

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2170. Min Act: -0.0001

Data Index: 1246301 (C4 (Web Text))

Max Activating Token Index: 931

Text #16

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2167. Min Act: -0.0001

Data Index: 237314 (C4 (Web Text))

Max Activating Token Index: 89

Text #17

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2085. Min Act: -0.0001

Data Index: 687557 (C4 (Web Text))

Max Activating Token Index: 42

Text #18

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2238. Min Act: -0.0001

Data Index: 1082285 (C4 (Web Text))

Max Activating Token Index: 853

Text #19

Max Range: 0.3022. Min Range: -0.3022

Max Act: 0.2207. Min Act: -0.0001

Data Index: 1309759 (C4 (Web Text))

Max Activating Token Index: 528