Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

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

Neuron 643 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l')

Text #0

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.3331. Min Act: -0.0001

Data Index: 649570 (C4 (Web Text))

Max Activating Token Index: 519

Text #1

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.3154. Min Act: -0.0001

Data Index: 1100992 (C4 (Web Text))

Max Activating Token Index: 568

Text #2

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.2347. Min Act: -0.0002

Data Index: 602038 (C4 (Web Text))

Max Activating Token Index: 996

Text #3

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.2226. Min Act: -0.0002

Data Index: 594775 (C4 (Web Text))

Max Activating Token Index: 389

Text #4

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.1649. Min Act: -0.0001

Data Index: 354752 (C4 (Web Text))

Max Activating Token Index: 990

Text #5

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.1743. Min Act: -0.0002

Data Index: 151947 (C4 (Web Text))

Max Activating Token Index: 619

Text #6

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.0861. Min Act: -0.0002

Data Index: 349141 (C4 (Web Text))

Max Activating Token Index: 863

Text #7

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.0396. Min Act: -0.0002

Data Index: 198242 (C4 (Web Text))

Max Activating Token Index: 332

Text #8

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.0360. Min Act: -0.0001

Data Index: 1462025 (Python Code)

Max Activating Token Index: 108

Text #9

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.1938. Min Act: -0.0002

Data Index: 348408 (C4 (Web Text))

Max Activating Token Index: 998

Text #10

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.0741. Min Act: -0.0001

Data Index: 954938 (C4 (Web Text))

Max Activating Token Index: 231

Text #11

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.1147. Min Act: -0.0002

Data Index: 178400 (C4 (Web Text))

Max Activating Token Index: 715

Text #12

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.2063. Min Act: -0.0002

Data Index: 472244 (C4 (Web Text))

Max Activating Token Index: 1020

Text #13

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.0753. Min Act: -0.0001

Data Index: 479790 (C4 (Web Text))

Max Activating Token Index: 490

Text #14

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.1348. Min Act: -0.0002

Data Index: 373830 (C4 (Web Text))

Max Activating Token Index: 76

Text #15

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.0988. Min Act: -0.0002

Data Index: 1289949 (C4 (Web Text))

Max Activating Token Index: 189

Text #16

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.0607. Min Act: -0.0001

Data Index: 1496808 (Python Code)

Max Activating Token Index: 190

Text #17

Max Range: 1.3331. Min Range: -1.3331

Max Act: 0.9659. Min Act: -0.0002

Data Index: 957480 (C4 (Web Text))

Max Activating Token Index: 598

Text #18

Max Range: 1.3331. Min Range: -1.3331

Max Act: 0.9868. Min Act: -0.0001

Data Index: 1179764 (C4 (Web Text))

Max Activating Token Index: 693

Text #19

Max Range: 1.3331. Min Range: -1.3331

Max Act: 1.0774. Min Act: -0.0002

Data Index: 837446 (C4 (Web Text))

Max Activating Token Index: 1000