Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 4787 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.4150. Min Act: -0.0000

Data Index: 1491510 (Python Code)

Max Activating Token Index: 912

Text #1

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.3555. Min Act: -0.0000

Data Index: 1113673 (C4 (Web Text))

Max Activating Token Index: 998

Text #2

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.3355. Min Act: -0.0000

Data Index: 1034169 (C4 (Web Text))

Max Activating Token Index: 377

Text #3

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.3154. Min Act: -0.0000

Data Index: 316944 (C4 (Web Text))

Max Activating Token Index: 909

Text #4

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.3287. Min Act: -0.0000

Data Index: 516929 (C4 (Web Text))

Max Activating Token Index: 24

Text #5

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.3328. Min Act: -0.0000

Data Index: 1321395 (C4 (Web Text))

Max Activating Token Index: 872

Text #6

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.3224. Min Act: -0.0000

Data Index: 640492 (C4 (Web Text))

Max Activating Token Index: 774

Text #7

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.3296. Min Act: -0.0000

Data Index: 543534 (C4 (Web Text))

Max Activating Token Index: 964

Text #8

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.3200. Min Act: -0.0000

Data Index: 338352 (C4 (Web Text))

Max Activating Token Index: 436

Text #9

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.3012. Min Act: -0.0000

Data Index: 131216 (C4 (Web Text))

Max Activating Token Index: 982

Text #10

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.3027. Min Act: -0.0000

Data Index: 632882 (C4 (Web Text))

Max Activating Token Index: 936

Text #11

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.2612. Min Act: -0.0000

Data Index: 1635342 (Python Code)

Max Activating Token Index: 258

Text #12

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.2819. Min Act: -0.0000

Data Index: 309971 (C4 (Web Text))

Max Activating Token Index: 459

Text #13

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.2908. Min Act: -0.0000

Data Index: 1306167 (C4 (Web Text))

Max Activating Token Index: 683

Text #14

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.3022. Min Act: -0.0000

Data Index: 321808 (C4 (Web Text))

Max Activating Token Index: 159

Text #15

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.2708. Min Act: -0.0000

Data Index: 1067600 (C4 (Web Text))

Max Activating Token Index: 956

Text #16

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.2852. Min Act: -0.0000

Data Index: 1337473 (C4 (Web Text))

Max Activating Token Index: 600

Text #17

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.2861. Min Act: -0.0000

Data Index: 919601 (C4 (Web Text))

Max Activating Token Index: 514

Text #18

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.2871. Min Act: -0.0000

Data Index: 1315107 (C4 (Web Text))

Max Activating Token Index: 1015

Text #19

Max Range: 0.4150. Min Range: -0.4150

Max Act: 0.2851. Min Act: -0.0000

Data Index: 908609 (C4 (Web Text))

Max Activating Token Index: 484