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

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

Neuron 3435 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 4.8672. Min Range: -4.8672

Max Act: 4.8672. Min Act: -0.0000

Data Index: 656655 (C4 (Web Text))

Max Activating Token Index: 710

Text #1

Max Range: 4.8672. Min Range: -4.8672

Max Act: 4.6285. Min Act: -0.0000

Data Index: 1105503 (C4 (Web Text))

Max Activating Token Index: 47

Text #2

Max Range: 4.8672. Min Range: -4.8672

Max Act: 4.4217. Min Act: -0.0000

Data Index: 270207 (C4 (Web Text))

Max Activating Token Index: 5

Text #3

Max Range: 4.8672. Min Range: -4.8672

Max Act: 4.3093. Min Act: -0.0000

Data Index: 907140 (C4 (Web Text))

Max Activating Token Index: 785

Text #4

Max Range: 4.8672. Min Range: -4.8672

Max Act: 4.4635. Min Act: -0.0000

Data Index: 1058647 (C4 (Web Text))

Max Activating Token Index: 61

Text #5

Max Range: 4.8672. Min Range: -4.8672

Max Act: 4.2925. Min Act: -0.0000

Data Index: 1017128 (C4 (Web Text))

Max Activating Token Index: 574

Text #6

Max Range: 4.8672. Min Range: -4.8672

Max Act: 4.1045. Min Act: -0.0000

Data Index: 305130 (C4 (Web Text))

Max Activating Token Index: 81

Text #7

Max Range: 4.8672. Min Range: -4.8672

Max Act: 4.6280. Min Act: -0.0000

Data Index: 843439 (C4 (Web Text))

Max Activating Token Index: 757

Text #8

Max Range: 4.8672. Min Range: -4.8672

Max Act: 4.1486. Min Act: -0.0000

Data Index: 435246 (C4 (Web Text))

Max Activating Token Index: 298

Text #9

Max Range: 4.8672. Min Range: -4.8672

Max Act: 3.9903. Min Act: -0.0000

Data Index: 712553 (C4 (Web Text))

Max Activating Token Index: 922

Text #10

Max Range: 4.8672. Min Range: -4.8672

Max Act: 3.5364. Min Act: -0.0000

Data Index: 1198086 (C4 (Web Text))

Max Activating Token Index: 183

Text #11

Max Range: 4.8672. Min Range: -4.8672

Max Act: 3.6717. Min Act: -0.0000

Data Index: 644342 (C4 (Web Text))

Max Activating Token Index: 746

Text #12

Max Range: 4.8672. Min Range: -4.8672

Max Act: 3.7394. Min Act: -0.0000

Data Index: 979059 (C4 (Web Text))

Max Activating Token Index: 13

Text #13

Max Range: 4.8672. Min Range: -4.8672

Max Act: 3.4124. Min Act: -0.0000

Data Index: 462465 (C4 (Web Text))

Max Activating Token Index: 564

Text #14

Max Range: 4.8672. Min Range: -4.8672

Max Act: 3.3680. Min Act: -0.0000

Data Index: 1023163 (C4 (Web Text))

Max Activating Token Index: 665

Text #15

Max Range: 4.8672. Min Range: -4.8672

Max Act: 3.4855. Min Act: -0.0000

Data Index: 1445398 (Python Code)

Max Activating Token Index: 915

Text #16

Max Range: 4.8672. Min Range: -4.8672

Max Act: 3.1547. Min Act: -0.0000

Data Index: 1229937 (C4 (Web Text))

Max Activating Token Index: 119

Text #17

Max Range: 4.8672. Min Range: -4.8672

Max Act: 3.2684. Min Act: -0.0000

Data Index: 277301 (C4 (Web Text))

Max Activating Token Index: 43

Text #18

Max Range: 4.8672. Min Range: -4.8672

Max Act: 3.4154. Min Act: -0.0000

Data Index: 951377 (C4 (Web Text))

Max Activating Token Index: 877

Text #19

Max Range: 4.8672. Min Range: -4.8672

Max Act: 3.2671. Min Act: -0.0000

Data Index: 1108048 (C4 (Web Text))

Max Activating Token Index: 433