Model: SoLU Model: 10 Layers, 5120 Neurons per Layer

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

Neuron 4444 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 2.3751. Min Range: -2.3751

Max Act: 2.2747. Min Act: -0.0000

Data Index: 1500858 (Python Code)

Max Activating Token Index: 616

Text #1

Max Range: 2.3751. Min Range: -2.3751

Max Act: 2.1459. Min Act: -0.0000

Data Index: 1639320 (Python Code)

Max Activating Token Index: 933

Text #2

Max Range: 2.3751. Min Range: -2.3751

Max Act: 2.3751. Min Act: -0.0000

Data Index: 1585218 (Python Code)

Max Activating Token Index: 671

Text #3

Max Range: 2.3751. Min Range: -2.3751

Max Act: 2.2653. Min Act: -0.0000

Data Index: 1437701 (Python Code)

Max Activating Token Index: 635

Text #4

Max Range: 2.3751. Min Range: -2.3751

Max Act: 2.0705. Min Act: -0.0000

Data Index: 767696 (C4 (Web Text))

Max Activating Token Index: 926

Text #5

Max Range: 2.3751. Min Range: -2.3751

Max Act: 2.1777. Min Act: -0.0000

Data Index: 1554876 (Python Code)

Max Activating Token Index: 568

Text #6

Max Range: 2.3751. Min Range: -2.3751

Max Act: 2.1274. Min Act: -0.0000

Data Index: 1621181 (Python Code)

Max Activating Token Index: 119

Text #7

Max Range: 2.3751. Min Range: -2.3751

Max Act: 2.2263. Min Act: -0.0000

Data Index: 1543908 (Python Code)

Max Activating Token Index: 1021

Text #8

Max Range: 2.3751. Min Range: -2.3751

Max Act: 2.1115. Min Act: -0.0000

Data Index: 1627326 (Python Code)

Max Activating Token Index: 289

Text #9

Max Range: 2.3751. Min Range: -2.3751

Max Act: 2.1199. Min Act: -0.0000

Data Index: 1636915 (Python Code)

Max Activating Token Index: 653

Text #10

Max Range: 2.3751. Min Range: -2.3751

Max Act: 1.9336. Min Act: -0.0000

Data Index: 1524777 (Python Code)

Max Activating Token Index: 464

Text #11

Max Range: 2.3751. Min Range: -2.3751

Max Act: 1.9599. Min Act: -0.0000

Data Index: 1385278 (Python Code)

Max Activating Token Index: 226

Text #12

Max Range: 2.3751. Min Range: -2.3751

Max Act: 1.9482. Min Act: -0.0000

Data Index: 1515238 (Python Code)

Max Activating Token Index: 903

Text #13

Max Range: 2.3751. Min Range: -2.3751

Max Act: 1.9402. Min Act: -0.0000

Data Index: 1564889 (Python Code)

Max Activating Token Index: 339

Text #14

Max Range: 2.3751. Min Range: -2.3751

Max Act: 1.9339. Min Act: -0.0000

Data Index: 1448527 (Python Code)

Max Activating Token Index: 533

Text #15

Max Range: 2.3751. Min Range: -2.3751

Max Act: 1.8080. Min Act: -0.0000

Data Index: 432113 (C4 (Web Text))

Max Activating Token Index: 588

Text #16

Max Range: 2.3751. Min Range: -2.3751

Max Act: 1.9100. Min Act: -0.0000

Data Index: 1550999 (Python Code)

Max Activating Token Index: 474

Text #17

Max Range: 2.3751. Min Range: -2.3751

Max Act: 1.8899. Min Act: -0.0000

Data Index: 1577369 (Python Code)

Max Activating Token Index: 441

Text #18

Max Range: 2.3751. Min Range: -2.3751

Max Act: 1.8840. Min Act: -0.0000

Data Index: 1363152 (Python Code)

Max Activating Token Index: 147

Text #19

Max Range: 2.3751. Min Range: -2.3751

Max Act: 1.7763. Min Act: -0.0000

Data Index: 21073 (C4 (Web Text))

Max Activating Token Index: 351