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

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

Neuron 3243 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.1500. Min Range: -3.1500

Max Act: 3.1500. Min Act: -0.0000

Data Index: 842497 (C4 (Web Text))

Max Activating Token Index: 253

Text #1

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.8057. Min Act: -0.0000

Data Index: 749827 (C4 (Web Text))

Max Activating Token Index: 332

Text #2

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.7760. Min Act: -0.0000

Data Index: 798901 (C4 (Web Text))

Max Activating Token Index: 292

Text #3

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.6601. Min Act: -0.0000

Data Index: 1083297 (C4 (Web Text))

Max Activating Token Index: 587

Text #4

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.6006. Min Act: -0.0001

Data Index: 762021 (C4 (Web Text))

Max Activating Token Index: 918

Text #5

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.4614. Min Act: -0.0000

Data Index: 844587 (C4 (Web Text))

Max Activating Token Index: 489

Text #6

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.3676. Min Act: -0.0001

Data Index: 983418 (C4 (Web Text))

Max Activating Token Index: 108

Text #7

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.3236. Min Act: -0.0000

Data Index: 178832 (C4 (Web Text))

Max Activating Token Index: 1021

Text #8

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.0763. Min Act: -0.0001

Data Index: 304123 (C4 (Web Text))

Max Activating Token Index: 205

Text #9

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.2599. Min Act: -0.0001

Data Index: 59655 (C4 (Web Text))

Max Activating Token Index: 552

Text #10

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.2817. Min Act: -0.0000

Data Index: 811317 (C4 (Web Text))

Max Activating Token Index: 628

Text #11

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.2411. Min Act: -0.0000

Data Index: 480273 (C4 (Web Text))

Max Activating Token Index: 466

Text #12

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.2169. Min Act: -0.0001

Data Index: 1403746 (Python Code)

Max Activating Token Index: 360

Text #13

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.2317. Min Act: -0.0000

Data Index: 875816 (C4 (Web Text))

Max Activating Token Index: 704

Text #14

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.2357. Min Act: -0.0000

Data Index: 25222 (C4 (Web Text))

Max Activating Token Index: 938

Text #15

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.2043. Min Act: -0.0000

Data Index: 1156500 (C4 (Web Text))

Max Activating Token Index: 86

Text #16

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.0686. Min Act: -0.0001

Data Index: 271678 (C4 (Web Text))

Max Activating Token Index: 818

Text #17

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.1682. Min Act: -0.0000

Data Index: 272838 (C4 (Web Text))

Max Activating Token Index: 458

Text #18

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.1412. Min Act: -0.0001

Data Index: 1091795 (C4 (Web Text))

Max Activating Token Index: 279

Text #19

Max Range: 3.1500. Min Range: -3.1500

Max Act: 1.0683. Min Act: -0.0000

Data Index: 299883 (C4 (Web Text))

Max Activating Token Index: 125