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

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

Neuron 3275 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.6552. Min Act: -0.0000

Data Index: 613508 (C4 (Web Text))

Max Activating Token Index: 95

Text #1

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.6976. Min Act: -0.0000

Data Index: 1326524 (C4 (Web Text))

Max Activating Token Index: 598

Text #2

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.6364. Min Act: -0.0000

Data Index: 950710 (C4 (Web Text))

Max Activating Token Index: 128

Text #3

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.4953. Min Act: -0.0000

Data Index: 195172 (C4 (Web Text))

Max Activating Token Index: 612

Text #4

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.5490. Min Act: -0.0000

Data Index: 1307266 (C4 (Web Text))

Max Activating Token Index: 168

Text #5

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.6002. Min Act: -0.0000

Data Index: 406880 (C4 (Web Text))

Max Activating Token Index: 509

Text #6

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.4890. Min Act: -0.0000

Data Index: 644164 (C4 (Web Text))

Max Activating Token Index: 145

Text #7

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.4786. Min Act: -0.0000

Data Index: 150679 (C4 (Web Text))

Max Activating Token Index: 12

Text #8

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.4568. Min Act: -0.0000

Data Index: 1163263 (C4 (Web Text))

Max Activating Token Index: 65

Text #9

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.4541. Min Act: -0.0000

Data Index: 234400 (C4 (Web Text))

Max Activating Token Index: 377

Text #10

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.5197. Min Act: -0.0000

Data Index: 1131695 (C4 (Web Text))

Max Activating Token Index: 956

Text #11

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.4067. Min Act: -0.0000

Data Index: 965630 (C4 (Web Text))

Max Activating Token Index: 355

Text #12

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.4323. Min Act: -0.0000

Data Index: 834990 (C4 (Web Text))

Max Activating Token Index: 636

Text #13

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.4437. Min Act: -0.0000

Data Index: 1331583 (C4 (Web Text))

Max Activating Token Index: 715

Text #14

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.4018. Min Act: -0.0000

Data Index: 1394185 (Python Code)

Max Activating Token Index: 513

Text #15

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.4411. Min Act: -0.0000

Data Index: 992982 (C4 (Web Text))

Max Activating Token Index: 1001

Text #16

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.4085. Min Act: -0.0000

Data Index: 306185 (C4 (Web Text))

Max Activating Token Index: 968

Text #17

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.4156. Min Act: -0.0000

Data Index: 18191 (C4 (Web Text))

Max Activating Token Index: 837

Text #18

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.3858. Min Act: -0.0000

Data Index: 1179583 (C4 (Web Text))

Max Activating Token Index: 94

Text #19

Max Range: 0.6976. Min Range: -0.6976

Max Act: 0.3929. Min Act: -0.0000

Data Index: 1291559 (C4 (Web Text))

Max Activating Token Index: 418