Model: SoLU Model: 6 Layers, 3072 Neurons per Layer

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

Neuron 403 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-6l')

Text #0

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.2239. Min Act: -0.0001

Data Index: 696414 (C4 (Web Text))

Max Activating Token Index: 968

Text #1

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1898. Min Act: -0.0001

Data Index: 1558533 (Python Code)

Max Activating Token Index: 401

Text #2

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1848. Min Act: -0.0001

Data Index: 1535041 (Python Code)

Max Activating Token Index: 258

Text #3

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1821. Min Act: -0.0001

Data Index: 850434 (C4 (Web Text))

Max Activating Token Index: 471

Text #4

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.2036. Min Act: -0.0001

Data Index: 1198777 (C4 (Web Text))

Max Activating Token Index: 580

Text #5

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1748. Min Act: -0.0001

Data Index: 1636370 (Python Code)

Max Activating Token Index: 694

Text #6

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.2007. Min Act: -0.0001

Data Index: 919771 (C4 (Web Text))

Max Activating Token Index: 644

Text #7

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1879. Min Act: -0.0001

Data Index: 320430 (C4 (Web Text))

Max Activating Token Index: 983

Text #8

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1834. Min Act: -0.0001

Data Index: 1340514 (C4 (Web Text))

Max Activating Token Index: 763

Text #9

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1741. Min Act: -0.0001

Data Index: 1368327 (Python Code)

Max Activating Token Index: 249

Text #10

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1719. Min Act: -0.0001

Data Index: 469486 (C4 (Web Text))

Max Activating Token Index: 340

Text #11

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1900. Min Act: -0.0001

Data Index: 435266 (C4 (Web Text))

Max Activating Token Index: 849

Text #12

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1589. Min Act: -0.0001

Data Index: 1572206 (Python Code)

Max Activating Token Index: 360

Text #13

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1632. Min Act: -0.0001

Data Index: 783390 (C4 (Web Text))

Max Activating Token Index: 130

Text #14

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1748. Min Act: -0.0001

Data Index: 623965 (C4 (Web Text))

Max Activating Token Index: 548

Text #15

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1610. Min Act: -0.0001

Data Index: 1152382 (C4 (Web Text))

Max Activating Token Index: 465

Text #16

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1625. Min Act: -0.0001

Data Index: 1474476 (Python Code)

Max Activating Token Index: 333

Text #17

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1491. Min Act: -0.0001

Data Index: 140935 (C4 (Web Text))

Max Activating Token Index: 295

Text #18

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1519. Min Act: -0.0001

Data Index: 217405 (C4 (Web Text))

Max Activating Token Index: 954

Text #19

Max Range: 0.2239. Min Range: -0.2239

Max Act: 0.1554. Min Act: -0.0001

Data Index: 301788 (C4 (Web Text))

Max Activating Token Index: 272