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

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

Neuron 3600 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l')

Text #0

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.8953. Min Act: -0.0000

Data Index: 1261240 (C4 (Web Text))

Max Activating Token Index: 933

Text #1

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.9031. Min Act: -0.0000

Data Index: 1217547 (C4 (Web Text))

Max Activating Token Index: 160

Text #2

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.8552. Min Act: -0.0000

Data Index: 1009639 (C4 (Web Text))

Max Activating Token Index: 836

Text #3

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.8447. Min Act: -0.0000

Data Index: 283 (C4 (Web Text))

Max Activating Token Index: 584

Text #4

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.8940. Min Act: -0.0000

Data Index: 589401 (C4 (Web Text))

Max Activating Token Index: 122

Text #5

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.8314. Min Act: -0.0000

Data Index: 1200873 (C4 (Web Text))

Max Activating Token Index: 945

Text #6

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.8291. Min Act: -0.0000

Data Index: 525044 (C4 (Web Text))

Max Activating Token Index: 552

Text #7

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.8428. Min Act: -0.0000

Data Index: 711286 (C4 (Web Text))

Max Activating Token Index: 78

Text #8

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.8497. Min Act: -0.0000

Data Index: 1137524 (C4 (Web Text))

Max Activating Token Index: 767

Text #9

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.8123. Min Act: -0.0000

Data Index: 635938 (C4 (Web Text))

Max Activating Token Index: 276

Text #10

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.7602. Min Act: -0.0000

Data Index: 101714 (C4 (Web Text))

Max Activating Token Index: 947

Text #11

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.7734. Min Act: -0.0000

Data Index: 39726 (C4 (Web Text))

Max Activating Token Index: 567

Text #12

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.8068. Min Act: -0.0001

Data Index: 1061339 (C4 (Web Text))

Max Activating Token Index: 185

Text #13

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.7771. Min Act: -0.0000

Data Index: 867880 (C4 (Web Text))

Max Activating Token Index: 918

Text #14

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.7747. Min Act: -0.0000

Data Index: 238294 (C4 (Web Text))

Max Activating Token Index: 685

Text #15

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.7677. Min Act: -0.0000

Data Index: 1202156 (C4 (Web Text))

Max Activating Token Index: 745

Text #16

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.7620. Min Act: -0.0000

Data Index: 990068 (C4 (Web Text))

Max Activating Token Index: 369

Text #17

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.7494. Min Act: -0.0000

Data Index: 475364 (C4 (Web Text))

Max Activating Token Index: 974

Text #18

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.7522. Min Act: -0.0000

Data Index: 1087347 (C4 (Web Text))

Max Activating Token Index: 592

Text #19

Max Range: 0.9031. Min Range: -0.9031

Max Act: 0.7170. Min Act: -0.0000

Data Index: 1122907 (C4 (Web Text))

Max Activating Token Index: 179