Model: SoLU Model: 6 Layers, 3072 Neurons per Layer

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

Neuron 535 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-6l')

Text #0

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.1807. Min Act: -0.0001

Data Index: 77910 (C4 (Web Text))

Max Activating Token Index: 399

Text #1

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.1387. Min Act: -0.0001

Data Index: 487712 (C4 (Web Text))

Max Activating Token Index: 10

Text #2

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.1206. Min Act: -0.0001

Data Index: 1420827 (Python Code)

Max Activating Token Index: 193

Text #3

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0996. Min Act: -0.0001

Data Index: 1218488 (C4 (Web Text))

Max Activating Token Index: 990

Text #4

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0849. Min Act: -0.0001

Data Index: 1208373 (C4 (Web Text))

Max Activating Token Index: 35

Text #5

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.1013. Min Act: -0.0001

Data Index: 843018 (C4 (Web Text))

Max Activating Token Index: 71

Text #6

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0924. Min Act: -0.0001

Data Index: 1609535 (Python Code)

Max Activating Token Index: 83

Text #7

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0922. Min Act: -0.0001

Data Index: 1483333 (Python Code)

Max Activating Token Index: 70

Text #8

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0886. Min Act: -0.0001

Data Index: 407561 (C4 (Web Text))

Max Activating Token Index: 685

Text #9

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0839. Min Act: -0.0001

Data Index: 1109594 (C4 (Web Text))

Max Activating Token Index: 272

Text #10

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0734. Min Act: -0.0001

Data Index: 130260 (C4 (Web Text))

Max Activating Token Index: 41

Text #11

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0738. Min Act: -0.0001

Data Index: 1606079 (Python Code)

Max Activating Token Index: 263

Text #12

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0666. Min Act: -0.0001

Data Index: 1023720 (C4 (Web Text))

Max Activating Token Index: 897

Text #13

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0672. Min Act: -0.0001

Data Index: 530146 (C4 (Web Text))

Max Activating Token Index: 943

Text #14

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0630. Min Act: -0.0001

Data Index: 997018 (C4 (Web Text))

Max Activating Token Index: 715

Text #15

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0696. Min Act: -0.0001

Data Index: 1379326 (Python Code)

Max Activating Token Index: 370

Text #16

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0687. Min Act: -0.0001

Data Index: 1257051 (C4 (Web Text))

Max Activating Token Index: 6

Text #17

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0703. Min Act: -0.0001

Data Index: 293453 (C4 (Web Text))

Max Activating Token Index: 101

Text #18

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0701. Min Act: -0.0001

Data Index: 1413633 (Python Code)

Max Activating Token Index: 828

Text #19

Max Range: 0.1807. Min Range: -0.1807

Max Act: 0.0704. Min Act: -0.0001

Data Index: 74591 (C4 (Web Text))

Max Activating Token Index: 548