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

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

Neuron 2708 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.6794. Min Act: -0.0000

Data Index: 1103225 (C4 (Web Text))

Max Activating Token Index: 319

Text #1

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.9410. Min Act: -0.0000

Data Index: 1117981 (C4 (Web Text))

Max Activating Token Index: 72

Text #2

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.4407. Min Act: -0.0000

Data Index: 810506 (C4 (Web Text))

Max Activating Token Index: 970

Text #3

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.2939. Min Act: -0.0000

Data Index: 39112 (C4 (Web Text))

Max Activating Token Index: 545

Text #4

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.1737. Min Act: -0.0000

Data Index: 1326293 (C4 (Web Text))

Max Activating Token Index: 431

Text #5

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.1899. Min Act: -0.0000

Data Index: 1588181 (Python Code)

Max Activating Token Index: 944

Text #6

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.2188. Min Act: -0.0000

Data Index: 831590 (C4 (Web Text))

Max Activating Token Index: 237

Text #7

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.0268. Min Act: -0.0000

Data Index: 471115 (C4 (Web Text))

Max Activating Token Index: 413

Text #8

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.0560. Min Act: -0.0000

Data Index: 1146619 (C4 (Web Text))

Max Activating Token Index: 179

Text #9

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.0546. Min Act: -0.0000

Data Index: 1370748 (Python Code)

Max Activating Token Index: 528

Text #10

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.0541. Min Act: -0.0000

Data Index: 703221 (C4 (Web Text))

Max Activating Token Index: 468

Text #11

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.0842. Min Act: -0.0000

Data Index: 1578792 (Python Code)

Max Activating Token Index: 673

Text #12

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.0827. Min Act: -0.0000

Data Index: 1182614 (C4 (Web Text))

Max Activating Token Index: 936

Text #13

Max Range: 1.9410. Min Range: -1.9410

Max Act: 0.9616. Min Act: -0.0000

Data Index: 1000632 (C4 (Web Text))

Max Activating Token Index: 692

Text #14

Max Range: 1.9410. Min Range: -1.9410

Max Act: 0.9743. Min Act: -0.0000

Data Index: 1369657 (Python Code)

Max Activating Token Index: 553

Text #15

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.0639. Min Act: -0.0000

Data Index: 1199404 (C4 (Web Text))

Max Activating Token Index: 306

Text #16

Max Range: 1.9410. Min Range: -1.9410

Max Act: 1.0302. Min Act: -0.0000

Data Index: 785947 (C4 (Web Text))

Max Activating Token Index: 192

Text #17

Max Range: 1.9410. Min Range: -1.9410

Max Act: 0.9669. Min Act: -0.0000

Data Index: 523330 (C4 (Web Text))

Max Activating Token Index: 967

Text #18

Max Range: 1.9410. Min Range: -1.9410

Max Act: 0.9762. Min Act: -0.0000

Data Index: 462209 (C4 (Web Text))

Max Activating Token Index: 787

Text #19

Max Range: 1.9410. Min Range: -1.9410

Max Act: 0.8112. Min Act: -0.0000

Data Index: 1556200 (Python Code)

Max Activating Token Index: 739