Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 3707 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.5732. Min Act: -0.0001

Data Index: 1573510 (Python Code)

Max Activating Token Index: 3

Text #1

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.5970. Min Act: -0.0001

Data Index: 45989 (C4 (Web Text))

Max Activating Token Index: 564

Text #2

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.6473. Min Act: -0.0001

Data Index: 1653768 (Python Code)

Max Activating Token Index: 886

Text #3

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4646. Min Act: -0.0001

Data Index: 364528 (C4 (Web Text))

Max Activating Token Index: 3

Text #4

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4848. Min Act: -0.0001

Data Index: 1388898 (Python Code)

Max Activating Token Index: 940

Text #5

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4679. Min Act: -0.0001

Data Index: 1524428 (Python Code)

Max Activating Token Index: 372

Text #6

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4616. Min Act: -0.0001

Data Index: 1401981 (Python Code)

Max Activating Token Index: 929

Text #7

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4074. Min Act: -0.0001

Data Index: 1236322 (C4 (Web Text))

Max Activating Token Index: 224

Text #8

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4412. Min Act: -0.0001

Data Index: 1508527 (Python Code)

Max Activating Token Index: 43

Text #9

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4115. Min Act: -0.0001

Data Index: 588532 (C4 (Web Text))

Max Activating Token Index: 330

Text #10

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4003. Min Act: -0.0001

Data Index: 549395 (C4 (Web Text))

Max Activating Token Index: 647

Text #11

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4834. Min Act: -0.0001

Data Index: 262979 (C4 (Web Text))

Max Activating Token Index: 764

Text #12

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.3809. Min Act: -0.0001

Data Index: 15234 (C4 (Web Text))

Max Activating Token Index: 866

Text #13

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4192. Min Act: -0.0001

Data Index: 1527170 (Python Code)

Max Activating Token Index: 9

Text #14

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4536. Min Act: -0.0001

Data Index: 616732 (C4 (Web Text))

Max Activating Token Index: 207

Text #15

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4346. Min Act: -0.0001

Data Index: 296590 (C4 (Web Text))

Max Activating Token Index: 315

Text #16

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.4315. Min Act: -0.0001

Data Index: 712481 (C4 (Web Text))

Max Activating Token Index: 1014

Text #17

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.3777. Min Act: -0.0001

Data Index: 723404 (C4 (Web Text))

Max Activating Token Index: 384

Text #18

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.3850. Min Act: -0.0001

Data Index: 1614000 (Python Code)

Max Activating Token Index: 541

Text #19

Max Range: 0.6473. Min Range: -0.6473

Max Act: 0.3752. Min Act: -0.0001

Data Index: 1568502 (Python Code)

Max Activating Token Index: 787