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

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

Neuron 506 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: 1.3095. Min Range: -1.3095

Max Act: 1.2042. Min Act: -0.0000

Data Index: 1543785 (Python Code)

Max Activating Token Index: 957

Text #1

Max Range: 1.3095. Min Range: -1.3095

Max Act: 1.1401. Min Act: -0.0000

Data Index: 951095 (C4 (Web Text))

Max Activating Token Index: 777

Text #2

Max Range: 1.3095. Min Range: -1.3095

Max Act: 1.3095. Min Act: -0.0000

Data Index: 1592002 (Python Code)

Max Activating Token Index: 1012

Text #3

Max Range: 1.3095. Min Range: -1.3095

Max Act: 1.2149. Min Act: -0.0000

Data Index: 1616773 (Python Code)

Max Activating Token Index: 96

Text #4

Max Range: 1.3095. Min Range: -1.3095

Max Act: 1.1666. Min Act: -0.0000

Data Index: 751181 (C4 (Web Text))

Max Activating Token Index: 88

Text #5

Max Range: 1.3095. Min Range: -1.3095

Max Act: 1.1013. Min Act: -0.0000

Data Index: 838212 (C4 (Web Text))

Max Activating Token Index: 901

Text #6

Max Range: 1.3095. Min Range: -1.3095

Max Act: 1.0824. Min Act: -0.0000

Data Index: 338240 (C4 (Web Text))

Max Activating Token Index: 656

Text #7

Max Range: 1.3095. Min Range: -1.3095

Max Act: 1.1231. Min Act: -0.0000

Data Index: 1473460 (Python Code)

Max Activating Token Index: 204

Text #8

Max Range: 1.3095. Min Range: -1.3095

Max Act: 1.0411. Min Act: -0.0000

Data Index: 921116 (C4 (Web Text))

Max Activating Token Index: 710

Text #9

Max Range: 1.3095. Min Range: -1.3095

Max Act: 1.0409. Min Act: -0.0000

Data Index: 703914 (C4 (Web Text))

Max Activating Token Index: 250

Text #10

Max Range: 1.3095. Min Range: -1.3095

Max Act: 1.0064. Min Act: -0.0000

Data Index: 956569 (C4 (Web Text))

Max Activating Token Index: 559

Text #11

Max Range: 1.3095. Min Range: -1.3095

Max Act: 0.8299. Min Act: -0.0000

Data Index: 1610082 (Python Code)

Max Activating Token Index: 838

Text #12

Max Range: 1.3095. Min Range: -1.3095

Max Act: 0.9500. Min Act: -0.0000

Data Index: 1569366 (Python Code)

Max Activating Token Index: 890

Text #13

Max Range: 1.3095. Min Range: -1.3095

Max Act: 1.0122. Min Act: -0.0000

Data Index: 1589818 (Python Code)

Max Activating Token Index: 295

Text #14

Max Range: 1.3095. Min Range: -1.3095

Max Act: 0.9423. Min Act: -0.0000

Data Index: 601426 (C4 (Web Text))

Max Activating Token Index: 160

Text #15

Max Range: 1.3095. Min Range: -1.3095

Max Act: 0.9199. Min Act: -0.0000

Data Index: 1475588 (Python Code)

Max Activating Token Index: 214

Text #16

Max Range: 1.3095. Min Range: -1.3095

Max Act: 0.9100. Min Act: -0.0000

Data Index: 89315 (C4 (Web Text))

Max Activating Token Index: 336

Text #17

Max Range: 1.3095. Min Range: -1.3095

Max Act: 0.9828. Min Act: -0.0000

Data Index: 838588 (C4 (Web Text))

Max Activating Token Index: 75

Text #18

Max Range: 1.3095. Min Range: -1.3095

Max Act: 1.0010. Min Act: -0.0000

Data Index: 379594 (C4 (Web Text))

Max Activating Token Index: 509

Text #19

Max Range: 1.3095. Min Range: -1.3095

Max Act: 0.9420. Min Act: -0.0000

Data Index: 553451 (C4 (Web Text))

Max Activating Token Index: 874