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

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

Neuron 2638 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 2.7194. Min Range: -2.7194

Max Act: 2.7194. Min Act: -0.0000

Data Index: 1388911 (Python Code)

Max Activating Token Index: 984

Text #1

Max Range: 2.7194. Min Range: -2.7194

Max Act: 2.1883. Min Act: -0.0000

Data Index: 366090 (C4 (Web Text))

Max Activating Token Index: 81

Text #2

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.9499. Min Act: -0.0000

Data Index: 1115242 (C4 (Web Text))

Max Activating Token Index: 802

Text #3

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.8878. Min Act: -0.0000

Data Index: 1478292 (Python Code)

Max Activating Token Index: 454

Text #4

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.8983. Min Act: -0.0000

Data Index: 1650672 (Python Code)

Max Activating Token Index: 399

Text #5

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.7874. Min Act: -0.0000

Data Index: 342397 (C4 (Web Text))

Max Activating Token Index: 83

Text #6

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.8867. Min Act: -0.0000

Data Index: 1414662 (Python Code)

Max Activating Token Index: 505

Text #7

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.8739. Min Act: -0.0000

Data Index: 127396 (C4 (Web Text))

Max Activating Token Index: 526

Text #8

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.7684. Min Act: -0.0000

Data Index: 1122737 (C4 (Web Text))

Max Activating Token Index: 132

Text #9

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.7025. Min Act: -0.0000

Data Index: 919090 (C4 (Web Text))

Max Activating Token Index: 974

Text #10

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.6963. Min Act: -0.0000

Data Index: 169533 (C4 (Web Text))

Max Activating Token Index: 286

Text #11

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.7205. Min Act: -0.0000

Data Index: 1584347 (Python Code)

Max Activating Token Index: 227

Text #12

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.6748. Min Act: -0.0000

Data Index: 1359889 (Python Code)

Max Activating Token Index: 732

Text #13

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.6043. Min Act: -0.0000

Data Index: 130731 (C4 (Web Text))

Max Activating Token Index: 407

Text #14

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.6236. Min Act: -0.0000

Data Index: 710082 (C4 (Web Text))

Max Activating Token Index: 574

Text #15

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.6120. Min Act: -0.0000

Data Index: 828519 (C4 (Web Text))

Max Activating Token Index: 12

Text #16

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.6073. Min Act: -0.0000

Data Index: 462449 (C4 (Web Text))

Max Activating Token Index: 251

Text #17

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.5767. Min Act: -0.0000

Data Index: 1229121 (C4 (Web Text))

Max Activating Token Index: 474

Text #18

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.4749. Min Act: -0.0000

Data Index: 68836 (C4 (Web Text))

Max Activating Token Index: 677

Text #19

Max Range: 2.7194. Min Range: -2.7194

Max Act: 1.5798. Min Act: -0.0000

Data Index: 862708 (C4 (Web Text))

Max Activating Token Index: 941