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

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

Neuron 5604 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.1144. Min Range: -2.1144

Max Act: 2.0544. Min Act: -0.0000

Data Index: 1044673 (C4 (Web Text))

Max Activating Token Index: 834

Text #1

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.8998. Min Act: -0.0000

Data Index: 971988 (C4 (Web Text))

Max Activating Token Index: 992

Text #2

Max Range: 2.1144. Min Range: -2.1144

Max Act: 2.1144. Min Act: -0.0000

Data Index: 1084752 (C4 (Web Text))

Max Activating Token Index: 195

Text #3

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.9722. Min Act: -0.0000

Data Index: 335552 (C4 (Web Text))

Max Activating Token Index: 985

Text #4

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.8795. Min Act: -0.0000

Data Index: 616319 (C4 (Web Text))

Max Activating Token Index: 22

Text #5

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.8788. Min Act: -0.0000

Data Index: 30958 (C4 (Web Text))

Max Activating Token Index: 511

Text #6

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.9375. Min Act: -0.0000

Data Index: 1127858 (C4 (Web Text))

Max Activating Token Index: 92

Text #7

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.9305. Min Act: -0.0000

Data Index: 1103049 (C4 (Web Text))

Max Activating Token Index: 119

Text #8

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.8099. Min Act: -0.0000

Data Index: 944296 (C4 (Web Text))

Max Activating Token Index: 510

Text #9

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.8006. Min Act: -0.0000

Data Index: 764822 (C4 (Web Text))

Max Activating Token Index: 148

Text #10

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.9075. Min Act: -0.0000

Data Index: 1222456 (C4 (Web Text))

Max Activating Token Index: 98

Text #11

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.7793. Min Act: -0.0000

Data Index: 1211163 (C4 (Web Text))

Max Activating Token Index: 71

Text #12

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.7676. Min Act: -0.0000

Data Index: 1112101 (C4 (Web Text))

Max Activating Token Index: 127

Text #13

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.8624. Min Act: -0.0000

Data Index: 1617964 (Python Code)

Max Activating Token Index: 60

Text #14

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.7267. Min Act: -0.0000

Data Index: 319511 (C4 (Web Text))

Max Activating Token Index: 905

Text #15

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.7329. Min Act: -0.0000

Data Index: 1060404 (C4 (Web Text))

Max Activating Token Index: 693

Text #16

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.6248. Min Act: -0.0000

Data Index: 978204 (C4 (Web Text))

Max Activating Token Index: 46

Text #17

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.6821. Min Act: -0.0000

Data Index: 1120027 (C4 (Web Text))

Max Activating Token Index: 435

Text #18

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.5801. Min Act: -0.0000

Data Index: 565830 (C4 (Web Text))

Max Activating Token Index: 186

Text #19

Max Range: 2.1144. Min Range: -2.1144

Max Act: 1.6758. Min Act: -0.0000

Data Index: 1059453 (C4 (Web Text))

Max Activating Token Index: 432