Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 2628 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1896. Min Act: -0.0002

Data Index: 323630 (C4 (Web Text))

Max Activating Token Index: 831

Text #1

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1787. Min Act: -0.0002

Data Index: 633108 (C4 (Web Text))

Max Activating Token Index: 21

Text #2

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1879. Min Act: -0.0002

Data Index: 544863 (C4 (Web Text))

Max Activating Token Index: 13

Text #3

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1824. Min Act: -0.0002

Data Index: 1331527 (C4 (Web Text))

Max Activating Token Index: 726

Text #4

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1645. Min Act: -0.0001

Data Index: 680134 (C4 (Web Text))

Max Activating Token Index: 927

Text #5

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1520. Min Act: -0.0001

Data Index: 1107623 (C4 (Web Text))

Max Activating Token Index: 665

Text #6

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1666. Min Act: -0.0002

Data Index: 494296 (C4 (Web Text))

Max Activating Token Index: 400

Text #7

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1574. Min Act: -0.0001

Data Index: 1590267 (Python Code)

Max Activating Token Index: 376

Text #8

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1500. Min Act: -0.0001

Data Index: 1438925 (Python Code)

Max Activating Token Index: 830

Text #9

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1735. Min Act: -0.0002

Data Index: 247657 (C4 (Web Text))

Max Activating Token Index: 899

Text #10

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1482. Min Act: -0.0001

Data Index: 1381793 (Python Code)

Max Activating Token Index: 600

Text #11

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1523. Min Act: -0.0001

Data Index: 1400046 (Python Code)

Max Activating Token Index: 846

Text #12

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1660. Min Act: -0.0002

Data Index: 537818 (C4 (Web Text))

Max Activating Token Index: 581

Text #13

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1669. Min Act: -0.0001

Data Index: 1413070 (Python Code)

Max Activating Token Index: 742

Text #14

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1536. Min Act: -0.0001

Data Index: 1616077 (Python Code)

Max Activating Token Index: 316

Text #15

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1591. Min Act: -0.0001

Data Index: 607261 (C4 (Web Text))

Max Activating Token Index: 220

Text #16

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1484. Min Act: -0.0001

Data Index: 688872 (C4 (Web Text))

Max Activating Token Index: 167

Text #17

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1591. Min Act: -0.0001

Data Index: 1642255 (Python Code)

Max Activating Token Index: 459

Text #18

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1477. Min Act: -0.0001

Data Index: 1537022 (Python Code)

Max Activating Token Index: 972

Text #19

Max Range: 0.1896. Min Range: -0.1896

Max Act: 0.1475. Min Act: -0.0001

Data Index: 755501 (C4 (Web Text))

Max Activating Token Index: 230