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

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

Neuron 750 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: 0.0487. Min Range: -0.0487

Max Act: 0.0487. Min Act: -0.0001

Data Index: 881725 (C4 (Web Text))

Max Activating Token Index: 85

Text #1

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0394. Min Act: -0.0001

Data Index: 954008 (C4 (Web Text))

Max Activating Token Index: 667

Text #2

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0403. Min Act: -0.0001

Data Index: 1489167 (Python Code)

Max Activating Token Index: 343

Text #3

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0388. Min Act: -0.0001

Data Index: 614974 (C4 (Web Text))

Max Activating Token Index: 296

Text #4

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0372. Min Act: -0.0001

Data Index: 623406 (C4 (Web Text))

Max Activating Token Index: 325

Text #5

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0376. Min Act: -0.0001

Data Index: 165591 (C4 (Web Text))

Max Activating Token Index: 62

Text #6

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0355. Min Act: -0.0001

Data Index: 295869 (C4 (Web Text))

Max Activating Token Index: 887

Text #7

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0311. Min Act: -0.0001

Data Index: 143861 (C4 (Web Text))

Max Activating Token Index: 1001

Text #8

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0338. Min Act: -0.0001

Data Index: 1257034 (C4 (Web Text))

Max Activating Token Index: 1023

Text #9

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0342. Min Act: -0.0001

Data Index: 1084014 (C4 (Web Text))

Max Activating Token Index: 152

Text #10

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0334. Min Act: -0.0001

Data Index: 1518863 (Python Code)

Max Activating Token Index: 493

Text #11

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0302. Min Act: -0.0001

Data Index: 1172196 (C4 (Web Text))

Max Activating Token Index: 454

Text #12

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0332. Min Act: -0.0001

Data Index: 565056 (C4 (Web Text))

Max Activating Token Index: 177

Text #13

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0305. Min Act: -0.0001

Data Index: 1137934 (C4 (Web Text))

Max Activating Token Index: 935

Text #14

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0335. Min Act: -0.0001

Data Index: 1008273 (C4 (Web Text))

Max Activating Token Index: 621

Text #15

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0294. Min Act: -0.0001

Data Index: 1477763 (Python Code)

Max Activating Token Index: 244

Text #16

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0309. Min Act: -0.0001

Data Index: 941713 (C4 (Web Text))

Max Activating Token Index: 134

Text #17

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0286. Min Act: -0.0001

Data Index: 829904 (C4 (Web Text))

Max Activating Token Index: 426

Text #18

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0290. Min Act: -0.0001

Data Index: 106206 (C4 (Web Text))

Max Activating Token Index: 518

Text #19

Max Range: 0.0487. Min Range: -0.0487

Max Act: 0.0304. Min Act: -0.0001

Data Index: 279930 (C4 (Web Text))

Max Activating Token Index: 322