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

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

Neuron 72 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1521. Min Act: -0.0001

Data Index: 1190708 (C4 (Web Text))

Max Activating Token Index: 365

Text #1

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1528. Min Act: -0.0001

Data Index: 581295 (C4 (Web Text))

Max Activating Token Index: 527

Text #2

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1513. Min Act: -0.0001

Data Index: 568501 (C4 (Web Text))

Max Activating Token Index: 165

Text #3

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1341. Min Act: -0.0001

Data Index: 238862 (C4 (Web Text))

Max Activating Token Index: 287

Text #4

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1289. Min Act: -0.0001

Data Index: 619945 (C4 (Web Text))

Max Activating Token Index: 94

Text #5

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1432. Min Act: -0.0001

Data Index: 34428 (C4 (Web Text))

Max Activating Token Index: 35

Text #6

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1300. Min Act: -0.0001

Data Index: 422073 (C4 (Web Text))

Max Activating Token Index: 504

Text #7

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1405. Min Act: -0.0001

Data Index: 1242232 (C4 (Web Text))

Max Activating Token Index: 626

Text #8

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1305. Min Act: -0.0001

Data Index: 1052905 (C4 (Web Text))

Max Activating Token Index: 164

Text #9

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1269. Min Act: -0.0001

Data Index: 823439 (C4 (Web Text))

Max Activating Token Index: 35

Text #10

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1219. Min Act: -0.0001

Data Index: 852265 (C4 (Web Text))

Max Activating Token Index: 41

Text #11

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1302. Min Act: -0.0001

Data Index: 636413 (C4 (Web Text))

Max Activating Token Index: 98

Text #12

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1196. Min Act: -0.0001

Data Index: 924603 (C4 (Web Text))

Max Activating Token Index: 245

Text #13

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1131. Min Act: -0.0001

Data Index: 914960 (C4 (Web Text))

Max Activating Token Index: 277

Text #14

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1087. Min Act: -0.0001

Data Index: 620211 (C4 (Web Text))

Max Activating Token Index: 161

Text #15

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1113. Min Act: -0.0001

Data Index: 771792 (C4 (Web Text))

Max Activating Token Index: 104

Text #16

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1137. Min Act: -0.0001

Data Index: 659203 (C4 (Web Text))

Max Activating Token Index: 315

Text #17

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1150. Min Act: -0.0001

Data Index: 154408 (C4 (Web Text))

Max Activating Token Index: 237

Text #18

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1098. Min Act: -0.0001

Data Index: 1200293 (C4 (Web Text))

Max Activating Token Index: 52

Text #19

Max Range: 0.1528. Min Range: -0.1528

Max Act: 0.1132. Min Act: -0.0001

Data Index: 952468 (C4 (Web Text))

Max Activating Token Index: 8