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

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

Neuron 3858 in Layer 3

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.3797. Min Act: -0.0001

Data Index: 938820 (C4 (Web Text))

Max Activating Token Index: 567

Text #1

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.4227. Min Act: -0.0001

Data Index: 1004912 (C4 (Web Text))

Max Activating Token Index: 224

Text #2

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.2279. Min Act: -0.0001

Data Index: 985783 (C4 (Web Text))

Max Activating Token Index: 898

Text #3

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.2861. Min Act: -0.0001

Data Index: 472993 (C4 (Web Text))

Max Activating Token Index: 8

Text #4

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.2765. Min Act: -0.0001

Data Index: 875612 (C4 (Web Text))

Max Activating Token Index: 233

Text #5

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.2822. Min Act: -0.0001

Data Index: 1194832 (C4 (Web Text))

Max Activating Token Index: 15

Text #6

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.0866. Min Act: -0.0001

Data Index: 1281614 (C4 (Web Text))

Max Activating Token Index: 160

Text #7

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.1938. Min Act: -0.0001

Data Index: 631431 (C4 (Web Text))

Max Activating Token Index: 647

Text #8

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.1633. Min Act: -0.0001

Data Index: 651453 (C4 (Web Text))

Max Activating Token Index: 630

Text #9

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.1251. Min Act: -0.0001

Data Index: 275125 (C4 (Web Text))

Max Activating Token Index: 534

Text #10

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.1887. Min Act: -0.0001

Data Index: 1344566 (C4 (Web Text))

Max Activating Token Index: 42

Text #11

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.2135. Min Act: -0.0001

Data Index: 759399 (C4 (Web Text))

Max Activating Token Index: 548

Text #12

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.2014. Min Act: -0.0001

Data Index: 1398216 (Python Code)

Max Activating Token Index: 612

Text #13

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.1678. Min Act: -0.0001

Data Index: 1288638 (C4 (Web Text))

Max Activating Token Index: 126

Text #14

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.1002. Min Act: -0.0001

Data Index: 608655 (C4 (Web Text))

Max Activating Token Index: 614

Text #15

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.1943. Min Act: -0.0001

Data Index: 1344821 (C4 (Web Text))

Max Activating Token Index: 594

Text #16

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.0567. Min Act: -0.0001

Data Index: 1154909 (C4 (Web Text))

Max Activating Token Index: 416

Text #17

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.0819. Min Act: -0.0001

Data Index: 846788 (C4 (Web Text))

Max Activating Token Index: 62

Text #18

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.1172. Min Act: -0.0001

Data Index: 93068 (C4 (Web Text))

Max Activating Token Index: 19

Text #19

Max Range: 1.4227. Min Range: -1.4227

Max Act: 1.0929. Min Act: -0.0001

Data Index: 596438 (C4 (Web Text))

Max Activating Token Index: 67