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

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

Neuron 2500 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.8516. Min Range: -5.8516

Max Act: 5.8516. Min Act: -0.0000

Data Index: 224649 (C4 (Web Text))

Max Activating Token Index: 77

Text #1

Max Range: 5.8516. Min Range: -5.8516

Max Act: 5.5065. Min Act: -0.0000

Data Index: 517101 (C4 (Web Text))

Max Activating Token Index: 142

Text #2

Max Range: 5.8516. Min Range: -5.8516

Max Act: 5.2700. Min Act: -0.0000

Data Index: 753681 (C4 (Web Text))

Max Activating Token Index: 829

Text #3

Max Range: 5.8516. Min Range: -5.8516

Max Act: 5.3506. Min Act: -0.0000

Data Index: 384035 (C4 (Web Text))

Max Activating Token Index: 955

Text #4

Max Range: 5.8516. Min Range: -5.8516

Max Act: 5.1020. Min Act: -0.0000

Data Index: 2632 (C4 (Web Text))

Max Activating Token Index: 990

Text #5

Max Range: 5.8516. Min Range: -5.8516

Max Act: 5.2363. Min Act: -0.0000

Data Index: 613722 (C4 (Web Text))

Max Activating Token Index: 107

Text #6

Max Range: 5.8516. Min Range: -5.8516

Max Act: 5.0529. Min Act: -0.0000

Data Index: 1098050 (C4 (Web Text))

Max Activating Token Index: 636

Text #7

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.7238. Min Act: -0.0000

Data Index: 65975 (C4 (Web Text))

Max Activating Token Index: 780

Text #8

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.8641. Min Act: -0.0000

Data Index: 482232 (C4 (Web Text))

Max Activating Token Index: 420

Text #9

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.8313. Min Act: -0.0000

Data Index: 1050695 (C4 (Web Text))

Max Activating Token Index: 821

Text #10

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.7955. Min Act: -0.0000

Data Index: 397803 (C4 (Web Text))

Max Activating Token Index: 693

Text #11

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.8054. Min Act: -0.0000

Data Index: 274127 (C4 (Web Text))

Max Activating Token Index: 639

Text #12

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.5018. Min Act: -0.0000

Data Index: 23776 (C4 (Web Text))

Max Activating Token Index: 202

Text #13

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.6492. Min Act: -0.0000

Data Index: 617078 (C4 (Web Text))

Max Activating Token Index: 817

Text #14

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.4493. Min Act: -0.0000

Data Index: 344998 (C4 (Web Text))

Max Activating Token Index: 178

Text #15

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.5577. Min Act: -0.0000

Data Index: 1353827 (C4 (Web Text))

Max Activating Token Index: 878

Text #16

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.1893. Min Act: -0.0000

Data Index: 678981 (C4 (Web Text))

Max Activating Token Index: 938

Text #17

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.3218. Min Act: -0.0000

Data Index: 1099545 (C4 (Web Text))

Max Activating Token Index: 29

Text #18

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.3041. Min Act: -0.0000

Data Index: 138769 (C4 (Web Text))

Max Activating Token Index: 827

Text #19

Max Range: 5.8516. Min Range: -5.8516

Max Act: 4.2046. Min Act: -0.0000

Data Index: 85602 (C4 (Web Text))

Max Activating Token Index: 579