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

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

Neuron 3123 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.3517. Min Range: -3.3517

Max Act: 3.3517. Min Act: -0.0001

Data Index: 1372660 (Python Code)

Max Activating Token Index: 746

Text #1

Max Range: 3.3517. Min Range: -3.3517

Max Act: 3.1705. Min Act: -0.0001

Data Index: 1307853 (C4 (Web Text))

Max Activating Token Index: 794

Text #2

Max Range: 3.3517. Min Range: -3.3517

Max Act: 3.2318. Min Act: -0.0001

Data Index: 1554136 (Python Code)

Max Activating Token Index: 653

Text #3

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.6383. Min Act: -0.0001

Data Index: 1441128 (Python Code)

Max Activating Token Index: 993

Text #4

Max Range: 3.3517. Min Range: -3.3517

Max Act: 3.0342. Min Act: -0.0001

Data Index: 1499844 (Python Code)

Max Activating Token Index: 796

Text #5

Max Range: 3.3517. Min Range: -3.3517

Max Act: 3.0449. Min Act: -0.0001

Data Index: 1623627 (Python Code)

Max Activating Token Index: 559

Text #6

Max Range: 3.3517. Min Range: -3.3517

Max Act: 3.0864. Min Act: -0.0001

Data Index: 1631382 (Python Code)

Max Activating Token Index: 673

Text #7

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.9634. Min Act: -0.0001

Data Index: 1526959 (Python Code)

Max Activating Token Index: 281

Text #8

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.6742. Min Act: -0.0001

Data Index: 1468954 (Python Code)

Max Activating Token Index: 504

Text #9

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.9721. Min Act: -0.0001

Data Index: 1440874 (Python Code)

Max Activating Token Index: 461

Text #10

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.8193. Min Act: -0.0001

Data Index: 736943 (C4 (Web Text))

Max Activating Token Index: 36

Text #11

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.9518. Min Act: -0.0001

Data Index: 1477802 (Python Code)

Max Activating Token Index: 938

Text #12

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.6807. Min Act: -0.0001

Data Index: 659778 (C4 (Web Text))

Max Activating Token Index: 186

Text #13

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.9143. Min Act: -0.0001

Data Index: 218673 (C4 (Web Text))

Max Activating Token Index: 835

Text #14

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.4952. Min Act: -0.0001

Data Index: 1409349 (Python Code)

Max Activating Token Index: 777

Text #15

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.8890. Min Act: -0.0001

Data Index: 1623050 (Python Code)

Max Activating Token Index: 596

Text #16

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.7610. Min Act: -0.0001

Data Index: 1499401 (Python Code)

Max Activating Token Index: 598

Text #17

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.8869. Min Act: -0.0001

Data Index: 825565 (C4 (Web Text))

Max Activating Token Index: 986

Text #18

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.7053. Min Act: -0.0001

Data Index: 1471818 (Python Code)

Max Activating Token Index: 168

Text #19

Max Range: 3.3517. Min Range: -3.3517

Max Act: 2.5788. Min Act: -0.0001

Data Index: 1452913 (Python Code)

Max Activating Token Index: 687