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

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

Neuron 3534 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: 0.8734. Min Range: -0.8734

Max Act: 0.8734. Min Act: -0.0000

Data Index: 942807 (C4 (Web Text))

Max Activating Token Index: 331

Text #1

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.7898. Min Act: -0.0001

Data Index: 148831 (C4 (Web Text))

Max Activating Token Index: 143

Text #2

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.6464. Min Act: -0.0001

Data Index: 1170342 (C4 (Web Text))

Max Activating Token Index: 194

Text #3

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.5777. Min Act: -0.0001

Data Index: 430808 (C4 (Web Text))

Max Activating Token Index: 877

Text #4

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.5271. Min Act: -0.0000

Data Index: 14331 (C4 (Web Text))

Max Activating Token Index: 550

Text #5

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.5051. Min Act: -0.0000

Data Index: 662666 (C4 (Web Text))

Max Activating Token Index: 817

Text #6

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4767. Min Act: -0.0001

Data Index: 762194 (C4 (Web Text))

Max Activating Token Index: 720

Text #7

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.5166. Min Act: -0.0001

Data Index: 961513 (C4 (Web Text))

Max Activating Token Index: 56

Text #8

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4751. Min Act: -0.0001

Data Index: 1287178 (C4 (Web Text))

Max Activating Token Index: 340

Text #9

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4757. Min Act: -0.0001

Data Index: 994323 (C4 (Web Text))

Max Activating Token Index: 70

Text #10

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4823. Min Act: -0.0000

Data Index: 345720 (C4 (Web Text))

Max Activating Token Index: 179

Text #11

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4610. Min Act: -0.0000

Data Index: 559232 (C4 (Web Text))

Max Activating Token Index: 639

Text #12

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4563. Min Act: -0.0001

Data Index: 90619 (C4 (Web Text))

Max Activating Token Index: 131

Text #13

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4617. Min Act: -0.0000

Data Index: 1158955 (C4 (Web Text))

Max Activating Token Index: 371

Text #14

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4838. Min Act: -0.0000

Data Index: 264139 (C4 (Web Text))

Max Activating Token Index: 704

Text #15

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4356. Min Act: -0.0000

Data Index: 407941 (C4 (Web Text))

Max Activating Token Index: 673

Text #16

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4524. Min Act: -0.0000

Data Index: 1512887 (Python Code)

Max Activating Token Index: 306

Text #17

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4635. Min Act: -0.0001

Data Index: 1099005 (C4 (Web Text))

Max Activating Token Index: 594

Text #18

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4894. Min Act: -0.0000

Data Index: 1246785 (C4 (Web Text))

Max Activating Token Index: 389

Text #19

Max Range: 0.8734. Min Range: -0.8734

Max Act: 0.4704. Min Act: -0.0000

Data Index: 852715 (C4 (Web Text))

Max Activating Token Index: 106