Model: SoLU Model: 6 Layers, 3072 Neurons per Layer

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

Neuron 2827 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-6l')

Text #0

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0928. Min Act: -0.0001

Data Index: 12042 (C4 (Web Text))

Max Activating Token Index: 15

Text #1

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0686. Min Act: -0.0001

Data Index: 1033921 (C4 (Web Text))

Max Activating Token Index: 805

Text #2

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0719. Min Act: -0.0001

Data Index: 1044533 (C4 (Web Text))

Max Activating Token Index: 301

Text #3

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0642. Min Act: -0.0001

Data Index: 408621 (C4 (Web Text))

Max Activating Token Index: 766

Text #4

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0660. Min Act: -0.0001

Data Index: 1212067 (C4 (Web Text))

Max Activating Token Index: 439

Text #5

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0667. Min Act: -0.0001

Data Index: 890609 (C4 (Web Text))

Max Activating Token Index: 796

Text #6

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0641. Min Act: -0.0001

Data Index: 967900 (C4 (Web Text))

Max Activating Token Index: 209

Text #7

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0645. Min Act: -0.0001

Data Index: 1346166 (C4 (Web Text))

Max Activating Token Index: 298

Text #8

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0630. Min Act: -0.0001

Data Index: 920113 (C4 (Web Text))

Max Activating Token Index: 520

Text #9

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0598. Min Act: -0.0001

Data Index: 38649 (C4 (Web Text))

Max Activating Token Index: 868

Text #10

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0675. Min Act: -0.0001

Data Index: 1010311 (C4 (Web Text))

Max Activating Token Index: 631

Text #11

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0585. Min Act: -0.0001

Data Index: 141365 (C4 (Web Text))

Max Activating Token Index: 321

Text #12

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0631. Min Act: -0.0001

Data Index: 1467578 (Python Code)

Max Activating Token Index: 462

Text #13

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0651. Min Act: -0.0001

Data Index: 394329 (C4 (Web Text))

Max Activating Token Index: 16

Text #14

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0621. Min Act: -0.0001

Data Index: 606663 (C4 (Web Text))

Max Activating Token Index: 340

Text #15

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0600. Min Act: -0.0001

Data Index: 335193 (C4 (Web Text))

Max Activating Token Index: 377

Text #16

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0581. Min Act: -0.0001

Data Index: 1320407 (C4 (Web Text))

Max Activating Token Index: 344

Text #17

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0604. Min Act: -0.0001

Data Index: 1497329 (Python Code)

Max Activating Token Index: 104

Text #18

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0595. Min Act: -0.0001

Data Index: 1532387 (Python Code)

Max Activating Token Index: 222

Text #19

Max Range: 0.0928. Min Range: -0.0928

Max Act: 0.0577. Min Act: -0.0001

Data Index: 1327629 (C4 (Web Text))

Max Activating Token Index: 914