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

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

Neuron 1055 in Layer 11

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.9338. Min Act: -0.0000

Data Index: 160796 (C4 (Web Text))

Max Activating Token Index: 55

Text #1

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.6284. Min Act: -0.0000

Data Index: 1472185 (Python Code)

Max Activating Token Index: 844

Text #2

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.8402. Min Act: -0.0000

Data Index: 1534725 (Python Code)

Max Activating Token Index: 604

Text #3

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.6425. Min Act: -0.0000

Data Index: 465791 (C4 (Web Text))

Max Activating Token Index: 373

Text #4

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.4932. Min Act: -0.0000

Data Index: 1278795 (C4 (Web Text))

Max Activating Token Index: 607

Text #5

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.3086. Min Act: -0.0000

Data Index: 810805 (C4 (Web Text))

Max Activating Token Index: 972

Text #6

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.2750. Min Act: -0.0000

Data Index: 418311 (C4 (Web Text))

Max Activating Token Index: 682

Text #7

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.2104. Min Act: -0.0000

Data Index: 621270 (C4 (Web Text))

Max Activating Token Index: 434

Text #8

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.1999. Min Act: -0.0000

Data Index: 1300951 (C4 (Web Text))

Max Activating Token Index: 438

Text #9

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.1940. Min Act: -0.0000

Data Index: 1639243 (Python Code)

Max Activating Token Index: 315

Text #10

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.1793. Min Act: -0.0000

Data Index: 546991 (C4 (Web Text))

Max Activating Token Index: 464

Text #11

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.2859. Min Act: -0.0000

Data Index: 475275 (C4 (Web Text))

Max Activating Token Index: 773

Text #12

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.2508. Min Act: -0.0000

Data Index: 1400056 (Python Code)

Max Activating Token Index: 444

Text #13

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.1790. Min Act: -0.0000

Data Index: 355473 (C4 (Web Text))

Max Activating Token Index: 946

Text #14

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.0883. Min Act: -0.0000

Data Index: 1530000 (Python Code)

Max Activating Token Index: 755

Text #15

Max Range: 2.9338. Min Range: -2.9338

Max Act: 1.9225. Min Act: -0.0000

Data Index: 1616123 (Python Code)

Max Activating Token Index: 841

Text #16

Max Range: 2.9338. Min Range: -2.9338

Max Act: 1.9998. Min Act: -0.0000

Data Index: 1325847 (C4 (Web Text))

Max Activating Token Index: 48

Text #17

Max Range: 2.9338. Min Range: -2.9338

Max Act: 1.9952. Min Act: -0.0000

Data Index: 1390991 (Python Code)

Max Activating Token Index: 319

Text #18

Max Range: 2.9338. Min Range: -2.9338

Max Act: 2.1080. Min Act: -0.0000

Data Index: 1118075 (C4 (Web Text))

Max Activating Token Index: 1022

Text #19

Max Range: 2.9338. Min Range: -2.9338

Max Act: 1.8954. Min Act: -0.0000

Data Index: 1523090 (Python Code)

Max Activating Token Index: 184