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

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

Neuron 1604 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.0967. Min Range: -3.0967

Max Act: 3.0967. Min Act: -0.0000

Data Index: 494921 (C4 (Web Text))

Max Activating Token Index: 198

Text #1

Max Range: 3.0967. Min Range: -3.0967

Max Act: 2.6767. Min Act: -0.0000

Data Index: 655003 (C4 (Web Text))

Max Activating Token Index: 300

Text #2

Max Range: 3.0967. Min Range: -3.0967

Max Act: 2.7233. Min Act: -0.0000

Data Index: 1337550 (C4 (Web Text))

Max Activating Token Index: 758

Text #3

Max Range: 3.0967. Min Range: -3.0967

Max Act: 2.3595. Min Act: -0.0001

Data Index: 1176276 (C4 (Web Text))

Max Activating Token Index: 895

Text #4

Max Range: 3.0967. Min Range: -3.0967

Max Act: 2.4221. Min Act: -0.0000

Data Index: 313073 (C4 (Web Text))

Max Activating Token Index: 901

Text #5

Max Range: 3.0967. Min Range: -3.0967

Max Act: 2.3083. Min Act: -0.0001

Data Index: 346743 (C4 (Web Text))

Max Activating Token Index: 402

Text #6

Max Range: 3.0967. Min Range: -3.0967

Max Act: 2.0850. Min Act: -0.0001

Data Index: 433855 (C4 (Web Text))

Max Activating Token Index: 383

Text #7

Max Range: 3.0967. Min Range: -3.0967

Max Act: 1.6340. Min Act: -0.0000

Data Index: 912388 (C4 (Web Text))

Max Activating Token Index: 130

Text #8

Max Range: 3.0967. Min Range: -3.0967

Max Act: 1.5822. Min Act: -0.0000

Data Index: 164590 (C4 (Web Text))

Max Activating Token Index: 455

Text #9

Max Range: 3.0967. Min Range: -3.0967

Max Act: 1.5752. Min Act: -0.0001

Data Index: 383421 (C4 (Web Text))

Max Activating Token Index: 213

Text #10

Max Range: 3.0967. Min Range: -3.0967

Max Act: 1.2431. Min Act: -0.0000

Data Index: 1080026 (C4 (Web Text))

Max Activating Token Index: 617

Text #11

Max Range: 3.0967. Min Range: -3.0967

Max Act: 1.5422. Min Act: -0.0001

Data Index: 887453 (C4 (Web Text))

Max Activating Token Index: 855

Text #12

Max Range: 3.0967. Min Range: -3.0967

Max Act: 1.1900. Min Act: -0.0000

Data Index: 773701 (C4 (Web Text))

Max Activating Token Index: 588

Text #13

Max Range: 3.0967. Min Range: -3.0967

Max Act: 1.0984. Min Act: -0.0000

Data Index: 1057085 (C4 (Web Text))

Max Activating Token Index: 243

Text #14

Max Range: 3.0967. Min Range: -3.0967

Max Act: 1.1946. Min Act: -0.0000

Data Index: 842890 (C4 (Web Text))

Max Activating Token Index: 858

Text #15

Max Range: 3.0967. Min Range: -3.0967

Max Act: 1.1147. Min Act: -0.0000

Data Index: 1007849 (C4 (Web Text))

Max Activating Token Index: 592

Text #16

Max Range: 3.0967. Min Range: -3.0967

Max Act: 1.0411. Min Act: -0.0001

Data Index: 1575003 (Python Code)

Max Activating Token Index: 104

Text #17

Max Range: 3.0967. Min Range: -3.0967

Max Act: 1.0318. Min Act: -0.0000

Data Index: 1553781 (Python Code)

Max Activating Token Index: 63

Text #18

Max Range: 3.0967. Min Range: -3.0967

Max Act: 1.0207. Min Act: -0.0000

Data Index: 265733 (C4 (Web Text))

Max Activating Token Index: 713

Text #19

Max Range: 3.0967. Min Range: -3.0967

Max Act: 0.9856. Min Act: -0.0000

Data Index: 1187870 (C4 (Web Text))

Max Activating Token Index: 612