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

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

Neuron 586 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: 0.8570. Min Range: -0.8570

Max Act: 0.8147. Min Act: -0.0000

Data Index: 1515455 (Python Code)

Max Activating Token Index: 90

Text #1

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.7860. Min Act: -0.0000

Data Index: 698082 (C4 (Web Text))

Max Activating Token Index: 390

Text #2

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.8570. Min Act: -0.0000

Data Index: 1603247 (Python Code)

Max Activating Token Index: 165

Text #3

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.7587. Min Act: -0.0000

Data Index: 1079194 (C4 (Web Text))

Max Activating Token Index: 25

Text #4

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.7534. Min Act: -0.0000

Data Index: 1372203 (Python Code)

Max Activating Token Index: 23

Text #5

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.7085. Min Act: -0.0001

Data Index: 1459384 (Python Code)

Max Activating Token Index: 747

Text #6

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.6740. Min Act: -0.0000

Data Index: 1361291 (Python Code)

Max Activating Token Index: 123

Text #7

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.6651. Min Act: -0.0001

Data Index: 80715 (C4 (Web Text))

Max Activating Token Index: 964

Text #8

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.7138. Min Act: -0.0001

Data Index: 424940 (C4 (Web Text))

Max Activating Token Index: 536

Text #9

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.6601. Min Act: -0.0000

Data Index: 1411026 (Python Code)

Max Activating Token Index: 788

Text #10

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.7769. Min Act: -0.0000

Data Index: 1094473 (C4 (Web Text))

Max Activating Token Index: 768

Text #11

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.6854. Min Act: -0.0000

Data Index: 1587513 (Python Code)

Max Activating Token Index: 301

Text #12

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.6300. Min Act: -0.0000

Data Index: 1552966 (Python Code)

Max Activating Token Index: 432

Text #13

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.6537. Min Act: -0.0000

Data Index: 997101 (C4 (Web Text))

Max Activating Token Index: 979

Text #14

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.6293. Min Act: -0.0000

Data Index: 960012 (C4 (Web Text))

Max Activating Token Index: 627

Text #15

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.5738. Min Act: -0.0000

Data Index: 1391419 (Python Code)

Max Activating Token Index: 1004

Text #16

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.6923. Min Act: -0.0001

Data Index: 971684 (C4 (Web Text))

Max Activating Token Index: 194

Text #17

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.5840. Min Act: -0.0000

Data Index: 1520236 (Python Code)

Max Activating Token Index: 362

Text #18

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.6197. Min Act: -0.0000

Data Index: 802793 (C4 (Web Text))

Max Activating Token Index: 1012

Text #19

Max Range: 0.8570. Min Range: -0.8570

Max Act: 0.5449. Min Act: -0.0000

Data Index: 1546589 (Python Code)

Max Activating Token Index: 278