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

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

Neuron 2817 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.3503. Min Range: -5.3503

Max Act: 5.3503. Min Act: 0.0000

Data Index: 1122657 (C4 (Web Text))

Max Activating Token Index: 557

Text #1

Max Range: 5.3503. Min Range: -5.3503

Max Act: 4.3501. Min Act: -0.0000

Data Index: 1242582 (C4 (Web Text))

Max Activating Token Index: 252

Text #2

Max Range: 5.3503. Min Range: -5.3503

Max Act: 4.3846. Min Act: -0.0000

Data Index: 1106114 (C4 (Web Text))

Max Activating Token Index: 895

Text #3

Max Range: 5.3503. Min Range: -5.3503

Max Act: 4.1971. Min Act: -0.0000

Data Index: 393312 (C4 (Web Text))

Max Activating Token Index: 488

Text #4

Max Range: 5.3503. Min Range: -5.3503

Max Act: 4.1510. Min Act: -0.0000

Data Index: 696414 (C4 (Web Text))

Max Activating Token Index: 785

Text #5

Max Range: 5.3503. Min Range: -5.3503

Max Act: 4.1262. Min Act: -0.0000

Data Index: 10460 (C4 (Web Text))

Max Activating Token Index: 967

Text #6

Max Range: 5.3503. Min Range: -5.3503

Max Act: 4.0520. Min Act: -0.0000

Data Index: 328232 (C4 (Web Text))

Max Activating Token Index: 624

Text #7

Max Range: 5.3503. Min Range: -5.3503

Max Act: 3.8092. Min Act: -0.0000

Data Index: 1585512 (Python Code)

Max Activating Token Index: 888

Text #8

Max Range: 5.3503. Min Range: -5.3503

Max Act: 3.9489. Min Act: 0.0000

Data Index: 1095423 (C4 (Web Text))

Max Activating Token Index: 316

Text #9

Max Range: 5.3503. Min Range: -5.3503

Max Act: 4.0817. Min Act: -0.0000

Data Index: 354241 (C4 (Web Text))

Max Activating Token Index: 968

Text #10

Max Range: 5.3503. Min Range: -5.3503

Max Act: 3.7288. Min Act: -0.0000

Data Index: 1593354 (Python Code)

Max Activating Token Index: 548

Text #11

Max Range: 5.3503. Min Range: -5.3503

Max Act: 3.8720. Min Act: -0.0000

Data Index: 1355636 (C4 (Web Text))

Max Activating Token Index: 590

Text #12

Max Range: 5.3503. Min Range: -5.3503

Max Act: 4.0185. Min Act: -0.0000

Data Index: 877306 (C4 (Web Text))

Max Activating Token Index: 614

Text #13

Max Range: 5.3503. Min Range: -5.3503

Max Act: 3.8048. Min Act: -0.0000

Data Index: 1368202 (Python Code)

Max Activating Token Index: 146

Text #14

Max Range: 5.3503. Min Range: -5.3503

Max Act: 3.7983. Min Act: -0.0000

Data Index: 1167358 (C4 (Web Text))

Max Activating Token Index: 681

Text #15

Max Range: 5.3503. Min Range: -5.3503

Max Act: 3.5454. Min Act: -0.0000

Data Index: 300954 (C4 (Web Text))

Max Activating Token Index: 718

Text #16

Max Range: 5.3503. Min Range: -5.3503

Max Act: 3.6541. Min Act: -0.0000

Data Index: 1328460 (C4 (Web Text))

Max Activating Token Index: 1001

Text #17

Max Range: 5.3503. Min Range: -5.3503

Max Act: 3.6001. Min Act: -0.0000

Data Index: 1546096 (Python Code)

Max Activating Token Index: 967

Text #18

Max Range: 5.3503. Min Range: -5.3503

Max Act: 3.5786. Min Act: -0.0000

Data Index: 1389316 (Python Code)

Max Activating Token Index: 138

Text #19

Max Range: 5.3503. Min Range: -5.3503

Max Act: 3.6869. Min Act: -0.0000

Data Index: 291197 (C4 (Web Text))

Max Activating Token Index: 532