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

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

Neuron 3201 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.7042. Min Act: -0.0000

Data Index: 1542127 (Python Code)

Max Activating Token Index: 12

Text #1

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.6590. Min Act: -0.0000

Data Index: 1130793 (C4 (Web Text))

Max Activating Token Index: 514

Text #2

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.6121. Min Act: -0.0000

Data Index: 376516 (C4 (Web Text))

Max Activating Token Index: 430

Text #3

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.6580. Min Act: -0.0000

Data Index: 994383 (C4 (Web Text))

Max Activating Token Index: 858

Text #4

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.5743. Min Act: -0.0000

Data Index: 1417641 (Python Code)

Max Activating Token Index: 501

Text #5

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.6252. Min Act: -0.0000

Data Index: 470260 (C4 (Web Text))

Max Activating Token Index: 793

Text #6

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.5969. Min Act: -0.0000

Data Index: 420608 (C4 (Web Text))

Max Activating Token Index: 760

Text #7

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.5747. Min Act: -0.0000

Data Index: 1229391 (C4 (Web Text))

Max Activating Token Index: 530

Text #8

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.4606. Min Act: -0.0000

Data Index: 1647500 (Python Code)

Max Activating Token Index: 858

Text #9

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.5326. Min Act: -0.0000

Data Index: 1401749 (Python Code)

Max Activating Token Index: 561

Text #10

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.4032. Min Act: -0.0000

Data Index: 1494613 (Python Code)

Max Activating Token Index: 819

Text #11

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.4486. Min Act: -0.0000

Data Index: 847296 (C4 (Web Text))

Max Activating Token Index: 322

Text #12

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.4947. Min Act: -0.0000

Data Index: 1016817 (C4 (Web Text))

Max Activating Token Index: 773

Text #13

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.4453. Min Act: -0.0000

Data Index: 1365156 (Python Code)

Max Activating Token Index: 119

Text #14

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.5180. Min Act: -0.0001

Data Index: 449041 (C4 (Web Text))

Max Activating Token Index: 437

Text #15

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.4643. Min Act: -0.0000

Data Index: 700881 (C4 (Web Text))

Max Activating Token Index: 439

Text #16

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.4644. Min Act: -0.0000

Data Index: 1490187 (Python Code)

Max Activating Token Index: 226

Text #17

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.4341. Min Act: -0.0000

Data Index: 334708 (C4 (Web Text))

Max Activating Token Index: 289

Text #18

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.4677. Min Act: -0.0000

Data Index: 811766 (C4 (Web Text))

Max Activating Token Index: 654

Text #19

Max Range: 0.7042. Min Range: -0.7042

Max Act: 0.4292. Min Act: -0.0001

Data Index: 304377 (C4 (Web Text))

Max Activating Token Index: 890