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

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

Neuron 4732 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: 3.8257. Min Range: -3.8257

Max Act: 3.8257. Min Act: -0.0001

Data Index: 879544 (C4 (Web Text))

Max Activating Token Index: 426

Text #1

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.2899. Min Act: -0.0001

Data Index: 1587304 (Python Code)

Max Activating Token Index: 855

Text #2

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.2415. Min Act: -0.0001

Data Index: 271574 (C4 (Web Text))

Max Activating Token Index: 215

Text #3

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.5319. Min Act: -0.0001

Data Index: 67834 (C4 (Web Text))

Max Activating Token Index: 962

Text #4

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.2212. Min Act: -0.0001

Data Index: 1113088 (C4 (Web Text))

Max Activating Token Index: 119

Text #5

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.2161. Min Act: -0.0001

Data Index: 1290071 (C4 (Web Text))

Max Activating Token Index: 32

Text #6

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.4075. Min Act: -0.0001

Data Index: 876324 (C4 (Web Text))

Max Activating Token Index: 726

Text #7

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.2549. Min Act: -0.0001

Data Index: 640677 (C4 (Web Text))

Max Activating Token Index: 469

Text #8

Max Range: 3.8257. Min Range: -3.8257

Max Act: 2.8391. Min Act: -0.0001

Data Index: 1277042 (C4 (Web Text))

Max Activating Token Index: 681

Text #9

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.0344. Min Act: -0.0001

Data Index: 707592 (C4 (Web Text))

Max Activating Token Index: 685

Text #10

Max Range: 3.8257. Min Range: -3.8257

Max Act: 2.7462. Min Act: -0.0001

Data Index: 1081844 (C4 (Web Text))

Max Activating Token Index: 238

Text #11

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.0420. Min Act: -0.0001

Data Index: 1359351 (C4 (Web Text))

Max Activating Token Index: 332

Text #12

Max Range: 3.8257. Min Range: -3.8257

Max Act: 2.7110. Min Act: -0.0001

Data Index: 1227270 (C4 (Web Text))

Max Activating Token Index: 773

Text #13

Max Range: 3.8257. Min Range: -3.8257

Max Act: 2.8987. Min Act: -0.0001

Data Index: 1404220 (Python Code)

Max Activating Token Index: 471

Text #14

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.1672. Min Act: -0.0001

Data Index: 1211673 (C4 (Web Text))

Max Activating Token Index: 856

Text #15

Max Range: 3.8257. Min Range: -3.8257

Max Act: 2.9117. Min Act: -0.0001

Data Index: 471735 (C4 (Web Text))

Max Activating Token Index: 309

Text #16

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.0442. Min Act: -0.0001

Data Index: 612748 (C4 (Web Text))

Max Activating Token Index: 836

Text #17

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.0236. Min Act: -0.0001

Data Index: 42438 (C4 (Web Text))

Max Activating Token Index: 1007

Text #18

Max Range: 3.8257. Min Range: -3.8257

Max Act: 2.7853. Min Act: -0.0001

Data Index: 169039 (C4 (Web Text))

Max Activating Token Index: 622

Text #19

Max Range: 3.8257. Min Range: -3.8257

Max Act: 3.0106. Min Act: -0.0001

Data Index: 293906 (C4 (Web Text))

Max Activating Token Index: 948