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

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

Neuron 4000 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.0325. Min Range: -1.0325

Max Act: 1.0109. Min Act: -0.0000

Data Index: 1378534 (Python Code)

Max Activating Token Index: 224

Text #1

Max Range: 1.0325. Min Range: -1.0325

Max Act: 1.0325. Min Act: -0.0000

Data Index: 922562 (C4 (Web Text))

Max Activating Token Index: 39

Text #2

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.8439. Min Act: -0.0000

Data Index: 891591 (C4 (Web Text))

Max Activating Token Index: 332

Text #3

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.8334. Min Act: -0.0000

Data Index: 183895 (C4 (Web Text))

Max Activating Token Index: 134

Text #4

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7831. Min Act: -0.0000

Data Index: 291528 (C4 (Web Text))

Max Activating Token Index: 64

Text #5

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7780. Min Act: -0.0000

Data Index: 315044 (C4 (Web Text))

Max Activating Token Index: 11

Text #6

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.8019. Min Act: -0.0000

Data Index: 1511889 (Python Code)

Max Activating Token Index: 430

Text #7

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7785. Min Act: -0.0000

Data Index: 1625947 (Python Code)

Max Activating Token Index: 841

Text #8

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7771. Min Act: -0.0000

Data Index: 883184 (C4 (Web Text))

Max Activating Token Index: 27

Text #9

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7621. Min Act: -0.0000

Data Index: 1639222 (Python Code)

Max Activating Token Index: 595

Text #10

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7851. Min Act: -0.0000

Data Index: 1267205 (C4 (Web Text))

Max Activating Token Index: 383

Text #11

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7593. Min Act: -0.0000

Data Index: 859021 (C4 (Web Text))

Max Activating Token Index: 23

Text #12

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7721. Min Act: -0.0000

Data Index: 921490 (C4 (Web Text))

Max Activating Token Index: 479

Text #13

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7028. Min Act: -0.0000

Data Index: 510643 (C4 (Web Text))

Max Activating Token Index: 694

Text #14

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7339. Min Act: -0.0000

Data Index: 703663 (C4 (Web Text))

Max Activating Token Index: 789

Text #15

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7380. Min Act: -0.0000

Data Index: 882118 (C4 (Web Text))

Max Activating Token Index: 47

Text #16

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7344. Min Act: -0.0000

Data Index: 149073 (C4 (Web Text))

Max Activating Token Index: 16

Text #17

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7014. Min Act: -0.0000

Data Index: 1018693 (C4 (Web Text))

Max Activating Token Index: 503

Text #18

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.6975. Min Act: -0.0000

Data Index: 861039 (C4 (Web Text))

Max Activating Token Index: 101

Text #19

Max Range: 1.0325. Min Range: -1.0325

Max Act: 0.7838. Min Act: -0.0000

Data Index: 319698 (C4 (Web Text))

Max Activating Token Index: 729