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

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

Neuron 906 in Layer 10

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.4182. Min Range: -2.4182

Max Act: 2.4182. Min Act: -0.0000

Data Index: 83143 (C4 (Web Text))

Max Activating Token Index: 861

Text #1

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.2789. Min Act: -0.0000

Data Index: 66200 (C4 (Web Text))

Max Activating Token Index: 161

Text #2

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.2752. Min Act: -0.0000

Data Index: 159551 (C4 (Web Text))

Max Activating Token Index: 864

Text #3

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.1978. Min Act: -0.0000

Data Index: 260055 (C4 (Web Text))

Max Activating Token Index: 326

Text #4

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.2220. Min Act: -0.0000

Data Index: 1625577 (Python Code)

Max Activating Token Index: 55

Text #5

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.1181. Min Act: -0.0000

Data Index: 818660 (C4 (Web Text))

Max Activating Token Index: 461

Text #6

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.1857. Min Act: -0.0000

Data Index: 814040 (C4 (Web Text))

Max Activating Token Index: 790

Text #7

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.0603. Min Act: -0.0000

Data Index: 1083536 (C4 (Web Text))

Max Activating Token Index: 315

Text #8

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.0330. Min Act: -0.0000

Data Index: 1013008 (C4 (Web Text))

Max Activating Token Index: 152

Text #9

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.0248. Min Act: -0.0000

Data Index: 53248 (C4 (Web Text))

Max Activating Token Index: 993

Text #10

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.0436. Min Act: -0.0000

Data Index: 816052 (C4 (Web Text))

Max Activating Token Index: 896

Text #11

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.0581. Min Act: 0.0000

Data Index: 152132 (C4 (Web Text))

Max Activating Token Index: 16

Text #12

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.0887. Min Act: -0.0000

Data Index: 561000 (C4 (Web Text))

Max Activating Token Index: 283

Text #13

Max Range: 2.4182. Min Range: -2.4182

Max Act: 0.9519. Min Act: 0.0000

Data Index: 589223 (C4 (Web Text))

Max Activating Token Index: 500

Text #14

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.0136. Min Act: 0.0000

Data Index: 347000 (C4 (Web Text))

Max Activating Token Index: 448

Text #15

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.0094. Min Act: -0.0000

Data Index: 56066 (C4 (Web Text))

Max Activating Token Index: 497

Text #16

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.0471. Min Act: -0.0000

Data Index: 790916 (C4 (Web Text))

Max Activating Token Index: 158

Text #17

Max Range: 2.4182. Min Range: -2.4182

Max Act: 1.0491. Min Act: -0.0000

Data Index: 52227 (C4 (Web Text))

Max Activating Token Index: 666

Text #18

Max Range: 2.4182. Min Range: -2.4182

Max Act: 0.9767. Min Act: -0.0000

Data Index: 656811 (C4 (Web Text))

Max Activating Token Index: 16

Text #19

Max Range: 2.4182. Min Range: -2.4182

Max Act: 0.9203. Min Act: -0.0000

Data Index: 589825 (C4 (Web Text))

Max Activating Token Index: 278