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

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

Neuron 4625 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: 3.1813. Min Range: -3.1813

Max Act: 3.1813. Min Act: -0.0000

Data Index: 1012051 (C4 (Web Text))

Max Activating Token Index: 404

Text #1

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.7255. Min Act: -0.0000

Data Index: 604925 (C4 (Web Text))

Max Activating Token Index: 990

Text #2

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.1877. Min Act: -0.0000

Data Index: 609308 (C4 (Web Text))

Max Activating Token Index: 709

Text #3

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.2878. Min Act: -0.0000

Data Index: 1373741 (Python Code)

Max Activating Token Index: 852

Text #4

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.4904. Min Act: -0.0000

Data Index: 1372191 (Python Code)

Max Activating Token Index: 233

Text #5

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.5005. Min Act: -0.0000

Data Index: 603293 (C4 (Web Text))

Max Activating Token Index: 1019

Text #6

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.2145. Min Act: -0.0000

Data Index: 1210467 (C4 (Web Text))

Max Activating Token Index: 964

Text #7

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.1665. Min Act: -0.0000

Data Index: 1449117 (Python Code)

Max Activating Token Index: 146

Text #8

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.1425. Min Act: -0.0000

Data Index: 81171 (C4 (Web Text))

Max Activating Token Index: 213

Text #9

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.0249. Min Act: -0.0000

Data Index: 1415316 (Python Code)

Max Activating Token Index: 823

Text #10

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.0970. Min Act: -0.0000

Data Index: 679639 (C4 (Web Text))

Max Activating Token Index: 404

Text #11

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.3128. Min Act: -0.0000

Data Index: 983028 (C4 (Web Text))

Max Activating Token Index: 1008

Text #12

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.1760. Min Act: -0.0000

Data Index: 46530 (C4 (Web Text))

Max Activating Token Index: 587

Text #13

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.2752. Min Act: -0.0000

Data Index: 132987 (C4 (Web Text))

Max Activating Token Index: 600

Text #14

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.2282. Min Act: -0.0000

Data Index: 1121818 (C4 (Web Text))

Max Activating Token Index: 288

Text #15

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.1053. Min Act: -0.0000

Data Index: 1420649 (Python Code)

Max Activating Token Index: 423

Text #16

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.0669. Min Act: -0.0000

Data Index: 543487 (C4 (Web Text))

Max Activating Token Index: 383

Text #17

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.2610. Min Act: -0.0000

Data Index: 1171410 (C4 (Web Text))

Max Activating Token Index: 714

Text #18

Max Range: 3.1813. Min Range: -3.1813

Max Act: 2.0170. Min Act: -0.0000

Data Index: 1504039 (Python Code)

Max Activating Token Index: 883

Text #19

Max Range: 3.1813. Min Range: -3.1813

Max Act: 1.8920. Min Act: -0.0000

Data Index: 433349 (C4 (Web Text))

Max Activating Token Index: 878