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

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

Neuron 3620 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2979. Min Act: -0.0001

Data Index: 247901 (C4 (Web Text))

Max Activating Token Index: 446

Text #1

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2873. Min Act: -0.0001

Data Index: 848960 (C4 (Web Text))

Max Activating Token Index: 92

Text #2

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2878. Min Act: -0.0001

Data Index: 149064 (C4 (Web Text))

Max Activating Token Index: 609

Text #3

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2677. Min Act: -0.0001

Data Index: 216989 (C4 (Web Text))

Max Activating Token Index: 889

Text #4

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2688. Min Act: -0.0001

Data Index: 772130 (C4 (Web Text))

Max Activating Token Index: 696

Text #5

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2597. Min Act: -0.0001

Data Index: 485956 (C4 (Web Text))

Max Activating Token Index: 66

Text #6

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2698. Min Act: -0.0001

Data Index: 325792 (C4 (Web Text))

Max Activating Token Index: 76

Text #7

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2723. Min Act: -0.0001

Data Index: 588392 (C4 (Web Text))

Max Activating Token Index: 291

Text #8

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2598. Min Act: -0.0001

Data Index: 530277 (C4 (Web Text))

Max Activating Token Index: 665

Text #9

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2678. Min Act: -0.0001

Data Index: 1241682 (C4 (Web Text))

Max Activating Token Index: 584

Text #10

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2726. Min Act: -0.0001

Data Index: 845567 (C4 (Web Text))

Max Activating Token Index: 558

Text #11

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2503. Min Act: -0.0001

Data Index: 125926 (C4 (Web Text))

Max Activating Token Index: 566

Text #12

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2583. Min Act: -0.0001

Data Index: 201485 (C4 (Web Text))

Max Activating Token Index: 38

Text #13

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2573. Min Act: -0.0001

Data Index: 91499 (C4 (Web Text))

Max Activating Token Index: 66

Text #14

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2649. Min Act: -0.0001

Data Index: 130780 (C4 (Web Text))

Max Activating Token Index: 183

Text #15

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2580. Min Act: -0.0001

Data Index: 1290959 (C4 (Web Text))

Max Activating Token Index: 589

Text #16

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2530. Min Act: -0.0001

Data Index: 156879 (C4 (Web Text))

Max Activating Token Index: 375

Text #17

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2610. Min Act: -0.0001

Data Index: 1202797 (C4 (Web Text))

Max Activating Token Index: 969

Text #18

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2418. Min Act: -0.0001

Data Index: 438863 (C4 (Web Text))

Max Activating Token Index: 718

Text #19

Max Range: 0.2979. Min Range: -0.2979

Max Act: 0.2414. Min Act: -0.0001

Data Index: 1247561 (C4 (Web Text))

Max Activating Token Index: 888