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

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

Neuron 2321 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.9023. Min Act: -0.0000

Data Index: 1520017 (Python Code)

Max Activating Token Index: 900

Text #1

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.8917. Min Act: -0.0000

Data Index: 1559191 (Python Code)

Max Activating Token Index: 611

Text #2

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.7801. Min Act: -0.0000

Data Index: 1492447 (Python Code)

Max Activating Token Index: 587

Text #3

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.8239. Min Act: -0.0000

Data Index: 1566949 (Python Code)

Max Activating Token Index: 646

Text #4

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.7860. Min Act: -0.0000

Data Index: 1510048 (Python Code)

Max Activating Token Index: 726

Text #5

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.7625. Min Act: -0.0000

Data Index: 232128 (C4 (Web Text))

Max Activating Token Index: 502

Text #6

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.7165. Min Act: -0.0000

Data Index: 1582812 (Python Code)

Max Activating Token Index: 813

Text #7

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.6772. Min Act: -0.0000

Data Index: 1515935 (Python Code)

Max Activating Token Index: 763

Text #8

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.7277. Min Act: -0.0000

Data Index: 879602 (C4 (Web Text))

Max Activating Token Index: 743

Text #9

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.6886. Min Act: -0.0000

Data Index: 1064649 (C4 (Web Text))

Max Activating Token Index: 93

Text #10

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.6828. Min Act: -0.0000

Data Index: 1652852 (Python Code)

Max Activating Token Index: 130

Text #11

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.6423. Min Act: -0.0000

Data Index: 1430687 (Python Code)

Max Activating Token Index: 284

Text #12

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.6421. Min Act: -0.0000

Data Index: 905142 (C4 (Web Text))

Max Activating Token Index: 461

Text #13

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.6578. Min Act: -0.0000

Data Index: 750059 (C4 (Web Text))

Max Activating Token Index: 42

Text #14

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.6325. Min Act: -0.0000

Data Index: 512517 (C4 (Web Text))

Max Activating Token Index: 22

Text #15

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.6162. Min Act: -0.0000

Data Index: 1579092 (Python Code)

Max Activating Token Index: 551

Text #16

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.6438. Min Act: -0.0000

Data Index: 1379494 (Python Code)

Max Activating Token Index: 896

Text #17

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.5984. Min Act: -0.0000

Data Index: 1653859 (Python Code)

Max Activating Token Index: 851

Text #18

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.5969. Min Act: -0.0000

Data Index: 831442 (C4 (Web Text))

Max Activating Token Index: 680

Text #19

Max Range: 0.9023. Min Range: -0.9023

Max Act: 0.5720. Min Act: -0.0000

Data Index: 721787 (C4 (Web Text))

Max Activating Token Index: 818