Model: SoLU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 630 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-3l')

Text #0

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0360. Min Act: -0.0003

Data Index: 552294 (C4 (Web Text))

Max Activating Token Index: 222

Text #1

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0355. Min Act: -0.0003

Data Index: 1235332 (C4 (Web Text))

Max Activating Token Index: 286

Text #2

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0334. Min Act: -0.0003

Data Index: 115324 (C4 (Web Text))

Max Activating Token Index: 280

Text #3

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0351. Min Act: -0.0003

Data Index: 1198910 (C4 (Web Text))

Max Activating Token Index: 589

Text #4

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0338. Min Act: -0.0003

Data Index: 670577 (C4 (Web Text))

Max Activating Token Index: 641

Text #5

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0325. Min Act: -0.0004

Data Index: 991448 (C4 (Web Text))

Max Activating Token Index: 777

Text #6

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0329. Min Act: -0.0003

Data Index: 1287170 (C4 (Web Text))

Max Activating Token Index: 378

Text #7

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0314. Min Act: -0.0003

Data Index: 1356831 (C4 (Web Text))

Max Activating Token Index: 266

Text #8

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0325. Min Act: -0.0004

Data Index: 466159 (C4 (Web Text))

Max Activating Token Index: 571

Text #9

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0325. Min Act: -0.0003

Data Index: 1019772 (C4 (Web Text))

Max Activating Token Index: 60

Text #10

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0299. Min Act: -0.0003

Data Index: 636440 (C4 (Web Text))

Max Activating Token Index: 250

Text #11

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0299. Min Act: -0.0003

Data Index: 1207640 (C4 (Web Text))

Max Activating Token Index: 860

Text #12

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0294. Min Act: -0.0003

Data Index: 202248 (C4 (Web Text))

Max Activating Token Index: 346

Text #13

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0305. Min Act: -0.0003

Data Index: 1485867 (Python Code)

Max Activating Token Index: 836

Text #14

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0292. Min Act: -0.0003

Data Index: 1500265 (Python Code)

Max Activating Token Index: 847

Text #15

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0300. Min Act: -0.0003

Data Index: 1250851 (C4 (Web Text))

Max Activating Token Index: 957

Text #16

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0300. Min Act: -0.0003

Data Index: 889053 (C4 (Web Text))

Max Activating Token Index: 781

Text #17

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0303. Min Act: -0.0004

Data Index: 285677 (C4 (Web Text))

Max Activating Token Index: 877

Text #18

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0304. Min Act: -0.0003

Data Index: 24097 (C4 (Web Text))

Max Activating Token Index: 342

Text #19

Max Range: 0.0360. Min Range: -0.0360

Max Act: 0.0289. Min Act: -0.0004

Data Index: 204225 (C4 (Web Text))

Max Activating Token Index: 457