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

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

Neuron 2287 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1309. Min Act: -0.0001

Data Index: 204287 (C4 (Web Text))

Max Activating Token Index: 293

Text #1

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1296. Min Act: -0.0001

Data Index: 1125561 (C4 (Web Text))

Max Activating Token Index: 317

Text #2

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1344. Min Act: -0.0001

Data Index: 601168 (C4 (Web Text))

Max Activating Token Index: 183

Text #3

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1190. Min Act: -0.0001

Data Index: 1148738 (C4 (Web Text))

Max Activating Token Index: 874

Text #4

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1315. Min Act: -0.0001

Data Index: 295362 (C4 (Web Text))

Max Activating Token Index: 5

Text #5

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1296. Min Act: -0.0001

Data Index: 1239580 (C4 (Web Text))

Max Activating Token Index: 174

Text #6

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1140. Min Act: -0.0001

Data Index: 1348 (C4 (Web Text))

Max Activating Token Index: 5

Text #7

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1054. Min Act: -0.0001

Data Index: 204677 (C4 (Web Text))

Max Activating Token Index: 81

Text #8

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1162. Min Act: -0.0001

Data Index: 884550 (C4 (Web Text))

Max Activating Token Index: 830

Text #9

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1053. Min Act: -0.0001

Data Index: 702506 (C4 (Web Text))

Max Activating Token Index: 981

Text #10

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1039. Min Act: -0.0001

Data Index: 486704 (C4 (Web Text))

Max Activating Token Index: 185

Text #11

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.0981. Min Act: -0.0001

Data Index: 1568691 (Python Code)

Max Activating Token Index: 966

Text #12

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1009. Min Act: -0.0001

Data Index: 42058 (C4 (Web Text))

Max Activating Token Index: 174

Text #13

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1087. Min Act: -0.0001

Data Index: 475571 (C4 (Web Text))

Max Activating Token Index: 383

Text #14

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1029. Min Act: -0.0001

Data Index: 1058253 (C4 (Web Text))

Max Activating Token Index: 767

Text #15

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.0994. Min Act: -0.0001

Data Index: 864878 (C4 (Web Text))

Max Activating Token Index: 4

Text #16

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.1023. Min Act: -0.0001

Data Index: 1181770 (C4 (Web Text))

Max Activating Token Index: 788

Text #17

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.0943. Min Act: -0.0001

Data Index: 1433138 (Python Code)

Max Activating Token Index: 195

Text #18

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.0830. Min Act: -0.0001

Data Index: 1075090 (C4 (Web Text))

Max Activating Token Index: 703

Text #19

Max Range: 0.1344. Min Range: -0.1344

Max Act: 0.0937. Min Act: -0.0001

Data Index: 1379548 (Python Code)

Max Activating Token Index: 313