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

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

Neuron 4763 in Layer 10

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.8690. Min Act: -0.0000

Data Index: 51521 (C4 (Web Text))

Max Activating Token Index: 529

Text #1

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.6700. Min Act: -0.0000

Data Index: 1121417 (C4 (Web Text))

Max Activating Token Index: 29

Text #2

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.1744. Min Act: -0.0000

Data Index: 1519437 (Python Code)

Max Activating Token Index: 167

Text #3

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.2649. Min Act: -0.0000

Data Index: 1528354 (Python Code)

Max Activating Token Index: 71

Text #4

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.2577. Min Act: -0.0000

Data Index: 616226 (C4 (Web Text))

Max Activating Token Index: 493

Text #5

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.2435. Min Act: -0.0000

Data Index: 589011 (C4 (Web Text))

Max Activating Token Index: 119

Text #6

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.3323. Min Act: -0.0000

Data Index: 309823 (C4 (Web Text))

Max Activating Token Index: 76

Text #7

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.0899. Min Act: -0.0000

Data Index: 213473 (C4 (Web Text))

Max Activating Token Index: 785

Text #8

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.1689. Min Act: -0.0000

Data Index: 485421 (C4 (Web Text))

Max Activating Token Index: 900

Text #9

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.1632. Min Act: -0.0000

Data Index: 1113307 (C4 (Web Text))

Max Activating Token Index: 424

Text #10

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.2468. Min Act: -0.0000

Data Index: 853190 (C4 (Web Text))

Max Activating Token Index: 766

Text #11

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.0569. Min Act: -0.0000

Data Index: 1462298 (Python Code)

Max Activating Token Index: 23

Text #12

Max Range: 2.8690. Min Range: -2.8690

Max Act: 1.9292. Min Act: -0.0000

Data Index: 510022 (C4 (Web Text))

Max Activating Token Index: 107

Text #13

Max Range: 2.8690. Min Range: -2.8690

Max Act: 2.0271. Min Act: -0.0000

Data Index: 1075790 (C4 (Web Text))

Max Activating Token Index: 624

Text #14

Max Range: 2.8690. Min Range: -2.8690

Max Act: 1.8954. Min Act: -0.0000

Data Index: 1493547 (Python Code)

Max Activating Token Index: 246

Text #15

Max Range: 2.8690. Min Range: -2.8690

Max Act: 1.9883. Min Act: -0.0000

Data Index: 1523236 (Python Code)

Max Activating Token Index: 222

Text #16

Max Range: 2.8690. Min Range: -2.8690

Max Act: 1.8932. Min Act: -0.0000

Data Index: 1454052 (Python Code)

Max Activating Token Index: 723

Text #17

Max Range: 2.8690. Min Range: -2.8690

Max Act: 1.7381. Min Act: -0.0000

Data Index: 1540687 (Python Code)

Max Activating Token Index: 32

Text #18

Max Range: 2.8690. Min Range: -2.8690

Max Act: 1.8316. Min Act: -0.0000

Data Index: 1527587 (Python Code)

Max Activating Token Index: 208

Text #19

Max Range: 2.8690. Min Range: -2.8690

Max Act: 1.7795. Min Act: -0.0000

Data Index: 649644 (C4 (Web Text))

Max Activating Token Index: 373