Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 512 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 1.1196. Min Range: -1.1196

Max Act: 1.1196. Min Act: -0.0001

Data Index: 1547217 (Python Code)

Max Activating Token Index: 94

Text #1

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6984. Min Act: -0.0000

Data Index: 1163994 (C4 (Web Text))

Max Activating Token Index: 679

Text #2

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6562. Min Act: -0.0001

Data Index: 1608110 (Python Code)

Max Activating Token Index: 556

Text #3

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6645. Min Act: -0.0001

Data Index: 190136 (C4 (Web Text))

Max Activating Token Index: 720

Text #4

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6841. Min Act: -0.0000

Data Index: 223634 (C4 (Web Text))

Max Activating Token Index: 986

Text #5

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6386. Min Act: -0.0001

Data Index: 450564 (C4 (Web Text))

Max Activating Token Index: 764

Text #6

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6588. Min Act: -0.0001

Data Index: 499331 (C4 (Web Text))

Max Activating Token Index: 26

Text #7

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6769. Min Act: -0.0001

Data Index: 1503574 (Python Code)

Max Activating Token Index: 62

Text #8

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6673. Min Act: -0.0001

Data Index: 1290946 (C4 (Web Text))

Max Activating Token Index: 989

Text #9

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6267. Min Act: -0.0001

Data Index: 1535474 (Python Code)

Max Activating Token Index: 650

Text #10

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.5818. Min Act: -0.0001

Data Index: 1391492 (Python Code)

Max Activating Token Index: 50

Text #11

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.5962. Min Act: -0.0001

Data Index: 1359877 (Python Code)

Max Activating Token Index: 449

Text #12

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6058. Min Act: -0.0001

Data Index: 1081910 (C4 (Web Text))

Max Activating Token Index: 452

Text #13

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6179. Min Act: -0.0000

Data Index: 291313 (C4 (Web Text))

Max Activating Token Index: 24

Text #14

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.5981. Min Act: -0.0001

Data Index: 1620251 (Python Code)

Max Activating Token Index: 190

Text #15

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.5958. Min Act: -0.0000

Data Index: 551679 (C4 (Web Text))

Max Activating Token Index: 900

Text #16

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6402. Min Act: -0.0001

Data Index: 1392060 (Python Code)

Max Activating Token Index: 294

Text #17

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.5270. Min Act: -0.0001

Data Index: 1567678 (Python Code)

Max Activating Token Index: 148

Text #18

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.6334. Min Act: -0.0001

Data Index: 1362664 (Python Code)

Max Activating Token Index: 105

Text #19

Max Range: 1.1196. Min Range: -1.1196

Max Act: 0.5696. Min Act: -0.0001

Data Index: 1365765 (Python Code)

Max Activating Token Index: 235