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

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

Neuron 2803 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.5161. Min Act: -0.0000

Data Index: 863652 (C4 (Web Text))

Max Activating Token Index: 222

Text #1

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.4599. Min Act: -0.0000

Data Index: 197875 (C4 (Web Text))

Max Activating Token Index: 944

Text #2

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.4349. Min Act: -0.0000

Data Index: 992469 (C4 (Web Text))

Max Activating Token Index: 57

Text #3

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.3728. Min Act: -0.0000

Data Index: 421934 (C4 (Web Text))

Max Activating Token Index: 30

Text #4

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.3261. Min Act: -0.0000

Data Index: 541522 (C4 (Web Text))

Max Activating Token Index: 686

Text #5

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.2984. Min Act: -0.0000

Data Index: 1597204 (Python Code)

Max Activating Token Index: 713

Text #6

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.3145. Min Act: -0.0000

Data Index: 1078237 (C4 (Web Text))

Max Activating Token Index: 305

Text #7

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.3106. Min Act: -0.0000

Data Index: 1257370 (C4 (Web Text))

Max Activating Token Index: 379

Text #8

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.3059. Min Act: -0.0000

Data Index: 1175164 (C4 (Web Text))

Max Activating Token Index: 857

Text #9

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.2882. Min Act: -0.0000

Data Index: 857165 (C4 (Web Text))

Max Activating Token Index: 722

Text #10

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.3063. Min Act: -0.0000

Data Index: 1196085 (C4 (Web Text))

Max Activating Token Index: 144

Text #11

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.3033. Min Act: -0.0000

Data Index: 183794 (C4 (Web Text))

Max Activating Token Index: 158

Text #12

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.2960. Min Act: -0.0000

Data Index: 1526597 (Python Code)

Max Activating Token Index: 687

Text #13

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.2866. Min Act: -0.0000

Data Index: 1600930 (Python Code)

Max Activating Token Index: 68

Text #14

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.2736. Min Act: -0.0000

Data Index: 1258644 (C4 (Web Text))

Max Activating Token Index: 757

Text #15

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.2762. Min Act: -0.0000

Data Index: 399488 (C4 (Web Text))

Max Activating Token Index: 175

Text #16

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.2660. Min Act: -0.0000

Data Index: 856893 (C4 (Web Text))

Max Activating Token Index: 693

Text #17

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.2830. Min Act: -0.0000

Data Index: 68435 (C4 (Web Text))

Max Activating Token Index: 68

Text #18

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.2781. Min Act: -0.0000

Data Index: 1176406 (C4 (Web Text))

Max Activating Token Index: 850

Text #19

Max Range: 0.5161. Min Range: -0.5161

Max Act: 0.2701. Min Act: -0.0000

Data Index: 56021 (C4 (Web Text))

Max Activating Token Index: 46