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

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

Neuron 4658 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: 0.6760. Min Range: -0.6760

Max Act: 0.6647. Min Act: -0.0000

Data Index: 538770 (C4 (Web Text))

Max Activating Token Index: 507

Text #1

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.6760. Min Act: -0.0000

Data Index: 876814 (C4 (Web Text))

Max Activating Token Index: 536

Text #2

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.5479. Min Act: -0.0000

Data Index: 1108795 (C4 (Web Text))

Max Activating Token Index: 597

Text #3

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.5419. Min Act: -0.0000

Data Index: 868932 (C4 (Web Text))

Max Activating Token Index: 738

Text #4

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.5330. Min Act: -0.0000

Data Index: 1374852 (Python Code)

Max Activating Token Index: 330

Text #5

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.5420. Min Act: -0.0000

Data Index: 588038 (C4 (Web Text))

Max Activating Token Index: 711

Text #6

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.5214. Min Act: -0.0000

Data Index: 282577 (C4 (Web Text))

Max Activating Token Index: 25

Text #7

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.5151. Min Act: -0.0000

Data Index: 969686 (C4 (Web Text))

Max Activating Token Index: 947

Text #8

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.5603. Min Act: -0.0000

Data Index: 774435 (C4 (Web Text))

Max Activating Token Index: 644

Text #9

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.5260. Min Act: -0.0000

Data Index: 981945 (C4 (Web Text))

Max Activating Token Index: 918

Text #10

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.5563. Min Act: -0.0000

Data Index: 148436 (C4 (Web Text))

Max Activating Token Index: 997

Text #11

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.4954. Min Act: -0.0000

Data Index: 429363 (C4 (Web Text))

Max Activating Token Index: 164

Text #12

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.4745. Min Act: -0.0000

Data Index: 1535779 (Python Code)

Max Activating Token Index: 26

Text #13

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.4896. Min Act: -0.0000

Data Index: 187472 (C4 (Web Text))

Max Activating Token Index: 1000

Text #14

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.4989. Min Act: -0.0000

Data Index: 1335687 (C4 (Web Text))

Max Activating Token Index: 842

Text #15

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.4729. Min Act: -0.0000

Data Index: 1614448 (Python Code)

Max Activating Token Index: 692

Text #16

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.4267. Min Act: -0.0000

Data Index: 92859 (C4 (Web Text))

Max Activating Token Index: 262

Text #17

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.4741. Min Act: -0.0000

Data Index: 351355 (C4 (Web Text))

Max Activating Token Index: 262

Text #18

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.4345. Min Act: -0.0000

Data Index: 188624 (C4 (Web Text))

Max Activating Token Index: 1001

Text #19

Max Range: 0.6760. Min Range: -0.6760

Max Act: 0.4615. Min Act: -0.0000

Data Index: 863525 (C4 (Web Text))

Max Activating Token Index: 678