Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 3682 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 5.2339. Min Range: -5.2339

Max Act: 5.2339. Min Act: -0.0001

Data Index: 83259 (C4 (Web Text))

Max Activating Token Index: 1001

Text #1

Max Range: 5.2339. Min Range: -5.2339

Max Act: 4.5422. Min Act: -0.0001

Data Index: 234067 (C4 (Web Text))

Max Activating Token Index: 990

Text #2

Max Range: 5.2339. Min Range: -5.2339

Max Act: 4.2189. Min Act: -0.0001

Data Index: 416928 (C4 (Web Text))

Max Activating Token Index: 996

Text #3

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.9967. Min Act: -0.0001

Data Index: 930361 (C4 (Web Text))

Max Activating Token Index: 881

Text #4

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.9245. Min Act: -0.0001

Data Index: 1350006 (C4 (Web Text))

Max Activating Token Index: 921

Text #5

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.9078. Min Act: -0.0001

Data Index: 548928 (C4 (Web Text))

Max Activating Token Index: 608

Text #6

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.6127. Min Act: -0.0001

Data Index: 227509 (C4 (Web Text))

Max Activating Token Index: 833

Text #7

Max Range: 5.2339. Min Range: -5.2339

Max Act: 4.0389. Min Act: -0.0001

Data Index: 989567 (C4 (Web Text))

Max Activating Token Index: 969

Text #8

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.4803. Min Act: -0.0001

Data Index: 287462 (C4 (Web Text))

Max Activating Token Index: 852

Text #9

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.9375. Min Act: -0.0001

Data Index: 1247689 (C4 (Web Text))

Max Activating Token Index: 875

Text #10

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.6463. Min Act: -0.0001

Data Index: 886263 (C4 (Web Text))

Max Activating Token Index: 350

Text #11

Max Range: 5.2339. Min Range: -5.2339

Max Act: 4.0714. Min Act: -0.0001

Data Index: 884017 (C4 (Web Text))

Max Activating Token Index: 139

Text #12

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.9058. Min Act: -0.0001

Data Index: 174708 (C4 (Web Text))

Max Activating Token Index: 788

Text #13

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.8822. Min Act: -0.0001

Data Index: 490190 (C4 (Web Text))

Max Activating Token Index: 431

Text #14

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.4455. Min Act: -0.0001

Data Index: 341562 (C4 (Web Text))

Max Activating Token Index: 249

Text #15

Max Range: 5.2339. Min Range: -5.2339

Max Act: 4.1549. Min Act: -0.0001

Data Index: 958864 (C4 (Web Text))

Max Activating Token Index: 544

Text #16

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.8091. Min Act: -0.0001

Data Index: 1112925 (C4 (Web Text))

Max Activating Token Index: 580

Text #17

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.6255. Min Act: -0.0001

Data Index: 223479 (C4 (Web Text))

Max Activating Token Index: 821

Text #18

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.5969. Min Act: -0.0001

Data Index: 98213 (C4 (Web Text))

Max Activating Token Index: 890

Text #19

Max Range: 5.2339. Min Range: -5.2339

Max Act: 3.6067. Min Act: -0.0001

Data Index: 654472 (C4 (Web Text))

Max Activating Token Index: 788