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

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

Neuron 534 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.3310. Min Range: -1.3310

Max Act: 1.3310. Min Act: -0.0001

Data Index: 349937 (C4 (Web Text))

Max Activating Token Index: 531

Text #1

Max Range: 1.3310. Min Range: -1.3310

Max Act: 1.1677. Min Act: -0.0001

Data Index: 876156 (C4 (Web Text))

Max Activating Token Index: 982

Text #2

Max Range: 1.3310. Min Range: -1.3310

Max Act: 1.0466. Min Act: -0.0001

Data Index: 103564 (C4 (Web Text))

Max Activating Token Index: 726

Text #3

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.9360. Min Act: -0.0001

Data Index: 1098756 (C4 (Web Text))

Max Activating Token Index: 785

Text #4

Max Range: 1.3310. Min Range: -1.3310

Max Act: 1.0136. Min Act: -0.0001

Data Index: 659668 (C4 (Web Text))

Max Activating Token Index: 250

Text #5

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.9893. Min Act: -0.0001

Data Index: 1071317 (C4 (Web Text))

Max Activating Token Index: 458

Text #6

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.9639. Min Act: -0.0001

Data Index: 495283 (C4 (Web Text))

Max Activating Token Index: 959

Text #7

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.9184. Min Act: -0.0000

Data Index: 1488384 (Python Code)

Max Activating Token Index: 93

Text #8

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.8621. Min Act: -0.0001

Data Index: 1133300 (C4 (Web Text))

Max Activating Token Index: 273

Text #9

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.9217. Min Act: -0.0001

Data Index: 927423 (C4 (Web Text))

Max Activating Token Index: 188

Text #10

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.9435. Min Act: -0.0001

Data Index: 1093605 (C4 (Web Text))

Max Activating Token Index: 760

Text #11

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.9158. Min Act: -0.0001

Data Index: 949872 (C4 (Web Text))

Max Activating Token Index: 348

Text #12

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.9199. Min Act: -0.0001

Data Index: 599854 (C4 (Web Text))

Max Activating Token Index: 813

Text #13

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.9350. Min Act: -0.0000

Data Index: 1592072 (Python Code)

Max Activating Token Index: 345

Text #14

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.8278. Min Act: -0.0001

Data Index: 146012 (C4 (Web Text))

Max Activating Token Index: 964

Text #15

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.8825. Min Act: -0.0001

Data Index: 1163187 (C4 (Web Text))

Max Activating Token Index: 679

Text #16

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.8481. Min Act: -0.0001

Data Index: 1409779 (Python Code)

Max Activating Token Index: 216

Text #17

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.8696. Min Act: -0.0000

Data Index: 1602452 (Python Code)

Max Activating Token Index: 243

Text #18

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.8184. Min Act: -0.0001

Data Index: 512405 (C4 (Web Text))

Max Activating Token Index: 204

Text #19

Max Range: 1.3310. Min Range: -1.3310

Max Act: 0.7227. Min Act: -0.0001

Data Index: 684139 (C4 (Web Text))

Max Activating Token Index: 400