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

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

Neuron 5572 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.7104. Min Act: -0.0000

Data Index: 1464580 (Python Code)

Max Activating Token Index: 443

Text #1

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.8094. Min Act: -0.0000

Data Index: 1634120 (Python Code)

Max Activating Token Index: 59

Text #2

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.4121. Min Act: -0.0001

Data Index: 1422598 (Python Code)

Max Activating Token Index: 864

Text #3

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.5679. Min Act: -0.0000

Data Index: 1199996 (C4 (Web Text))

Max Activating Token Index: 66

Text #4

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.6529. Min Act: -0.0000

Data Index: 312257 (C4 (Web Text))

Max Activating Token Index: 259

Text #5

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.6198. Min Act: -0.0001

Data Index: 1096848 (C4 (Web Text))

Max Activating Token Index: 750

Text #6

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.3119. Min Act: -0.0000

Data Index: 726402 (C4 (Web Text))

Max Activating Token Index: 123

Text #7

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.3713. Min Act: -0.0001

Data Index: 1621149 (Python Code)

Max Activating Token Index: 611

Text #8

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.2354. Min Act: -0.0000

Data Index: 513427 (C4 (Web Text))

Max Activating Token Index: 154

Text #9

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.3182. Min Act: -0.0000

Data Index: 79246 (C4 (Web Text))

Max Activating Token Index: 353

Text #10

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.0955. Min Act: -0.0000

Data Index: 392392 (C4 (Web Text))

Max Activating Token Index: 64

Text #11

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.3095. Min Act: -0.0001

Data Index: 606629 (C4 (Web Text))

Max Activating Token Index: 150

Text #12

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.1782. Min Act: -0.0000

Data Index: 292207 (C4 (Web Text))

Max Activating Token Index: 401

Text #13

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.1943. Min Act: -0.0000

Data Index: 964030 (C4 (Web Text))

Max Activating Token Index: 876

Text #14

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.1588. Min Act: -0.0000

Data Index: 1649240 (Python Code)

Max Activating Token Index: 113

Text #15

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.2419. Min Act: -0.0000

Data Index: 343312 (C4 (Web Text))

Max Activating Token Index: 125

Text #16

Max Range: 2.8094. Min Range: -2.8094

Max Act: 1.8925. Min Act: -0.0000

Data Index: 1312920 (C4 (Web Text))

Max Activating Token Index: 830

Text #17

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.0299. Min Act: -0.0000

Data Index: 712610 (C4 (Web Text))

Max Activating Token Index: 667

Text #18

Max Range: 2.8094. Min Range: -2.8094

Max Act: 2.4834. Min Act: -0.0000

Data Index: 1307596 (C4 (Web Text))

Max Activating Token Index: 150

Text #19

Max Range: 2.8094. Min Range: -2.8094

Max Act: 1.9076. Min Act: -0.0000

Data Index: 101526 (C4 (Web Text))

Max Activating Token Index: 220