Model: SoLU Model: 6 Layers, 3072 Neurons per Layer

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

Neuron 2353 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-6l')

Text #0

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3864. Min Act: -0.0003

Data Index: 463027 (C4 (Web Text))

Max Activating Token Index: 960

Text #1

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3604. Min Act: -0.0003

Data Index: 36476 (C4 (Web Text))

Max Activating Token Index: 235

Text #2

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3660. Min Act: -0.0003

Data Index: 834596 (C4 (Web Text))

Max Activating Token Index: 438

Text #3

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3671. Min Act: -0.0003

Data Index: 1087900 (C4 (Web Text))

Max Activating Token Index: 958

Text #4

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3419. Min Act: -0.0003

Data Index: 934264 (C4 (Web Text))

Max Activating Token Index: 306

Text #5

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3345. Min Act: -0.0003

Data Index: 486223 (C4 (Web Text))

Max Activating Token Index: 76

Text #6

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3231. Min Act: -0.0003

Data Index: 204708 (C4 (Web Text))

Max Activating Token Index: 331

Text #7

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3239. Min Act: -0.0003

Data Index: 994659 (C4 (Web Text))

Max Activating Token Index: 820

Text #8

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3224. Min Act: -0.0003

Data Index: 1632353 (Python Code)

Max Activating Token Index: 111

Text #9

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3001. Min Act: -0.0003

Data Index: 1165956 (C4 (Web Text))

Max Activating Token Index: 61

Text #10

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3354. Min Act: -0.0003

Data Index: 1023761 (C4 (Web Text))

Max Activating Token Index: 882

Text #11

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3171. Min Act: -0.0003

Data Index: 926300 (C4 (Web Text))

Max Activating Token Index: 574

Text #12

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3077. Min Act: -0.0003

Data Index: 343850 (C4 (Web Text))

Max Activating Token Index: 28

Text #13

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3034. Min Act: -0.0003

Data Index: 1072167 (C4 (Web Text))

Max Activating Token Index: 668

Text #14

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3127. Min Act: -0.0003

Data Index: 912570 (C4 (Web Text))

Max Activating Token Index: 984

Text #15

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3108. Min Act: -0.0003

Data Index: 234735 (C4 (Web Text))

Max Activating Token Index: 174

Text #16

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3023. Min Act: -0.0003

Data Index: 1001746 (C4 (Web Text))

Max Activating Token Index: 838

Text #17

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3195. Min Act: -0.0003

Data Index: 989931 (C4 (Web Text))

Max Activating Token Index: 447

Text #18

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3175. Min Act: -0.0003

Data Index: 22207 (C4 (Web Text))

Max Activating Token Index: 452

Text #19

Max Range: 0.3864. Min Range: -0.3864

Max Act: 0.3170. Min Act: -0.0003

Data Index: 1490110 (Python Code)

Max Activating Token Index: 587