Model: SoLU Model: 12 Layers, 6144 Neurons per Layer

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

Neuron 3707 in Layer 10

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-12l')

Text #0

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.4360. Min Act: -0.0000

Data Index: 178527 (C4 (Web Text))

Max Activating Token Index: 692

Text #1

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.4555. Min Act: -0.0000

Data Index: 294716 (C4 (Web Text))

Max Activating Token Index: 712

Text #2

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.4213. Min Act: -0.0000

Data Index: 322286 (C4 (Web Text))

Max Activating Token Index: 50

Text #3

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.4309. Min Act: 0.0000

Data Index: 59951 (C4 (Web Text))

Max Activating Token Index: 151

Text #4

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.3851. Min Act: 0.0000

Data Index: 939932 (C4 (Web Text))

Max Activating Token Index: 176

Text #5

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.3750. Min Act: 0.0000

Data Index: 824933 (C4 (Web Text))

Max Activating Token Index: 277

Text #6

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.3697. Min Act: 0.0000

Data Index: 139912 (C4 (Web Text))

Max Activating Token Index: 198

Text #7

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.3094. Min Act: -0.0000

Data Index: 1646710 (Python Code)

Max Activating Token Index: 801

Text #8

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.3271. Min Act: 0.0000

Data Index: 1157899 (C4 (Web Text))

Max Activating Token Index: 199

Text #9

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.3197. Min Act: -0.0000

Data Index: 1083973 (C4 (Web Text))

Max Activating Token Index: 518

Text #10

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.2833. Min Act: -0.0000

Data Index: 1456690 (Python Code)

Max Activating Token Index: 300

Text #11

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.2828. Min Act: -0.0000

Data Index: 1241566 (C4 (Web Text))

Max Activating Token Index: 806

Text #12

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.2891. Min Act: 0.0000

Data Index: 59366 (C4 (Web Text))

Max Activating Token Index: 756

Text #13

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.2753. Min Act: -0.0000

Data Index: 1552918 (Python Code)

Max Activating Token Index: 230

Text #14

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.2795. Min Act: 0.0001

Data Index: 641965 (C4 (Web Text))

Max Activating Token Index: 188

Text #15

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.3053. Min Act: -0.0000

Data Index: 15175 (C4 (Web Text))

Max Activating Token Index: 660

Text #16

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.2597. Min Act: 0.0000

Data Index: 567433 (C4 (Web Text))

Max Activating Token Index: 558

Text #17

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.2878. Min Act: 0.0000

Data Index: 807192 (C4 (Web Text))

Max Activating Token Index: 895

Text #18

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.2783. Min Act: 0.0000

Data Index: 756935 (C4 (Web Text))

Max Activating Token Index: 457

Text #19

Max Range: 0.4555. Min Range: -0.4555

Max Act: 0.2827. Min Act: -0.0000

Data Index: 885788 (C4 (Web Text))

Max Activating Token Index: 126