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

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

Neuron 4897 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: 0.6296. Min Range: -0.6296

Max Act: 0.5681. Min Act: -0.0001

Data Index: 471454 (C4 (Web Text))

Max Activating Token Index: 72

Text #1

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5501. Min Act: -0.0001

Data Index: 393610 (C4 (Web Text))

Max Activating Token Index: 16

Text #2

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.6296. Min Act: -0.0001

Data Index: 1289923 (C4 (Web Text))

Max Activating Token Index: 54

Text #3

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.6105. Min Act: -0.0001

Data Index: 371269 (C4 (Web Text))

Max Activating Token Index: 29

Text #4

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5683. Min Act: -0.0001

Data Index: 1376190 (Python Code)

Max Activating Token Index: 145

Text #5

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5592. Min Act: -0.0001

Data Index: 442836 (C4 (Web Text))

Max Activating Token Index: 23

Text #6

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5426. Min Act: -0.0001

Data Index: 267052 (C4 (Web Text))

Max Activating Token Index: 679

Text #7

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5335. Min Act: -0.0001

Data Index: 848695 (C4 (Web Text))

Max Activating Token Index: 69

Text #8

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5822. Min Act: -0.0001

Data Index: 211391 (C4 (Web Text))

Max Activating Token Index: 35

Text #9

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5303. Min Act: -0.0001

Data Index: 409401 (C4 (Web Text))

Max Activating Token Index: 135

Text #10

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5293. Min Act: -0.0001

Data Index: 1574721 (Python Code)

Max Activating Token Index: 15

Text #11

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5240. Min Act: -0.0001

Data Index: 1300686 (C4 (Web Text))

Max Activating Token Index: 675

Text #12

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5092. Min Act: -0.0001

Data Index: 69061 (C4 (Web Text))

Max Activating Token Index: 34

Text #13

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5347. Min Act: -0.0001

Data Index: 1044732 (C4 (Web Text))

Max Activating Token Index: 44

Text #14

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5163. Min Act: -0.0001

Data Index: 1087208 (C4 (Web Text))

Max Activating Token Index: 32

Text #15

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5229. Min Act: -0.0001

Data Index: 1069101 (C4 (Web Text))

Max Activating Token Index: 41

Text #16

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5217. Min Act: -0.0001

Data Index: 716299 (C4 (Web Text))

Max Activating Token Index: 291

Text #17

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5161. Min Act: -0.0001

Data Index: 1058471 (C4 (Web Text))

Max Activating Token Index: 815

Text #18

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5564. Min Act: -0.0001

Data Index: 1625774 (Python Code)

Max Activating Token Index: 971

Text #19

Max Range: 0.6296. Min Range: -0.6296

Max Act: 0.5206. Min Act: -0.0001

Data Index: 1117134 (C4 (Web Text))

Max Activating Token Index: 285