Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

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

Neuron 1968 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0988. Min Act: -0.0002

Data Index: 591147 (C4 (Web Text))

Max Activating Token Index: 831

Text #1

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0922. Min Act: -0.0002

Data Index: 823524 (C4 (Web Text))

Max Activating Token Index: 547

Text #2

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0985. Min Act: -0.0002

Data Index: 1094205 (C4 (Web Text))

Max Activating Token Index: 832

Text #3

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0896. Min Act: -0.0002

Data Index: 512105 (C4 (Web Text))

Max Activating Token Index: 657

Text #4

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0850. Min Act: -0.0002

Data Index: 422780 (C4 (Web Text))

Max Activating Token Index: 566

Text #5

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0786. Min Act: -0.0002

Data Index: 277728 (C4 (Web Text))

Max Activating Token Index: 813

Text #6

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0778. Min Act: -0.0002

Data Index: 345811 (C4 (Web Text))

Max Activating Token Index: 527

Text #7

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0758. Min Act: -0.0002

Data Index: 1150443 (C4 (Web Text))

Max Activating Token Index: 264

Text #8

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0737. Min Act: -0.0002

Data Index: 306668 (C4 (Web Text))

Max Activating Token Index: 104

Text #9

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0731. Min Act: -0.0002

Data Index: 546787 (C4 (Web Text))

Max Activating Token Index: 646

Text #10

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0676. Min Act: -0.0002

Data Index: 715598 (C4 (Web Text))

Max Activating Token Index: 257

Text #11

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0763. Min Act: -0.0002

Data Index: 1644262 (Python Code)

Max Activating Token Index: 341

Text #12

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0717. Min Act: -0.0002

Data Index: 963888 (C4 (Web Text))

Max Activating Token Index: 780

Text #13

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0689. Min Act: -0.0002

Data Index: 1009002 (C4 (Web Text))

Max Activating Token Index: 567

Text #14

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0687. Min Act: -0.0002

Data Index: 889432 (C4 (Web Text))

Max Activating Token Index: 500

Text #15

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0702. Min Act: -0.0002

Data Index: 644206 (C4 (Web Text))

Max Activating Token Index: 540

Text #16

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0725. Min Act: -0.0002

Data Index: 1249574 (C4 (Web Text))

Max Activating Token Index: 700

Text #17

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0694. Min Act: -0.0002

Data Index: 242378 (C4 (Web Text))

Max Activating Token Index: 1006

Text #18

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0685. Min Act: -0.0002

Data Index: 503355 (C4 (Web Text))

Max Activating Token Index: 936

Text #19

Max Range: 0.0988. Min Range: -0.0988

Max Act: 0.0691. Min Act: -0.0002

Data Index: 1011761 (C4 (Web Text))

Max Activating Token Index: 451