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

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

Neuron 395 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.6718. Min Range: -2.6718

Max Act: 2.6718. Min Act: -0.0001

Data Index: 529124 (C4 (Web Text))

Max Activating Token Index: 25

Text #1

Max Range: 2.6718. Min Range: -2.6718

Max Act: 2.3797. Min Act: -0.0000

Data Index: 1264911 (C4 (Web Text))

Max Activating Token Index: 193

Text #2

Max Range: 2.6718. Min Range: -2.6718

Max Act: 2.1752. Min Act: -0.0001

Data Index: 721570 (C4 (Web Text))

Max Activating Token Index: 994

Text #3

Max Range: 2.6718. Min Range: -2.6718

Max Act: 2.2356. Min Act: -0.0000

Data Index: 1164330 (C4 (Web Text))

Max Activating Token Index: 152

Text #4

Max Range: 2.6718. Min Range: -2.6718

Max Act: 2.0639. Min Act: -0.0000

Data Index: 1214233 (C4 (Web Text))

Max Activating Token Index: 156

Text #5

Max Range: 2.6718. Min Range: -2.6718

Max Act: 2.0199. Min Act: -0.0001

Data Index: 1492669 (Python Code)

Max Activating Token Index: 895

Text #6

Max Range: 2.6718. Min Range: -2.6718

Max Act: 2.0544. Min Act: -0.0000

Data Index: 373923 (C4 (Web Text))

Max Activating Token Index: 180

Text #7

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.8568. Min Act: -0.0000

Data Index: 360129 (C4 (Web Text))

Max Activating Token Index: 104

Text #8

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.9431. Min Act: -0.0000

Data Index: 277684 (C4 (Web Text))

Max Activating Token Index: 991

Text #9

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.8505. Min Act: -0.0000

Data Index: 479396 (C4 (Web Text))

Max Activating Token Index: 912

Text #10

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.6988. Min Act: 0.0000

Data Index: 23828 (C4 (Web Text))

Max Activating Token Index: 872

Text #11

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.7204. Min Act: -0.0001

Data Index: 168902 (C4 (Web Text))

Max Activating Token Index: 340

Text #12

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.7386. Min Act: -0.0000

Data Index: 2297 (C4 (Web Text))

Max Activating Token Index: 962

Text #13

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.7161. Min Act: -0.0000

Data Index: 1621235 (Python Code)

Max Activating Token Index: 47

Text #14

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.6397. Min Act: -0.0000

Data Index: 167413 (C4 (Web Text))

Max Activating Token Index: 81

Text #15

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.7247. Min Act: -0.0000

Data Index: 1093024 (C4 (Web Text))

Max Activating Token Index: 891

Text #16

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.6511. Min Act: -0.0001

Data Index: 717272 (C4 (Web Text))

Max Activating Token Index: 380

Text #17

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.6929. Min Act: 0.0000

Data Index: 1029159 (C4 (Web Text))

Max Activating Token Index: 841

Text #18

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.5243. Min Act: -0.0000

Data Index: 559498 (C4 (Web Text))

Max Activating Token Index: 411

Text #19

Max Range: 2.6718. Min Range: -2.6718

Max Act: 1.5556. Min Act: -0.0000

Data Index: 1273843 (C4 (Web Text))

Max Activating Token Index: 724