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

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

Neuron 4276 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: 5.0940. Min Range: -5.0940

Max Act: 4.8680. Min Act: -0.0000

Data Index: 629120 (C4 (Web Text))

Max Activating Token Index: 781

Text #1

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.9482. Min Act: -0.0000

Data Index: 963473 (C4 (Web Text))

Max Activating Token Index: 714

Text #2

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.8231. Min Act: -0.0001

Data Index: 469024 (C4 (Web Text))

Max Activating Token Index: 700

Text #3

Max Range: 5.0940. Min Range: -5.0940

Max Act: 5.0940. Min Act: -0.0000

Data Index: 658623 (C4 (Web Text))

Max Activating Token Index: 624

Text #4

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.8120. Min Act: -0.0000

Data Index: 1557816 (Python Code)

Max Activating Token Index: 998

Text #5

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.0609. Min Act: -0.0000

Data Index: 1286525 (C4 (Web Text))

Max Activating Token Index: 682

Text #6

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.6034. Min Act: -0.0000

Data Index: 1088524 (C4 (Web Text))

Max Activating Token Index: 983

Text #7

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.4161. Min Act: -0.0000

Data Index: 1608834 (Python Code)

Max Activating Token Index: 712

Text #8

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.6171. Min Act: -0.0000

Data Index: 747732 (C4 (Web Text))

Max Activating Token Index: 869

Text #9

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.2596. Min Act: -0.0000

Data Index: 179899 (C4 (Web Text))

Max Activating Token Index: 659

Text #10

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.3866. Min Act: -0.0000

Data Index: 790926 (C4 (Web Text))

Max Activating Token Index: 1000

Text #11

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.5430. Min Act: -0.0000

Data Index: 1583438 (Python Code)

Max Activating Token Index: 1006

Text #12

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.3529. Min Act: -0.0000

Data Index: 932017 (C4 (Web Text))

Max Activating Token Index: 438

Text #13

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.2909. Min Act: -0.0000

Data Index: 1498983 (Python Code)

Max Activating Token Index: 812

Text #14

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.2943. Min Act: -0.0000

Data Index: 502979 (C4 (Web Text))

Max Activating Token Index: 848

Text #15

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.1043. Min Act: -0.0000

Data Index: 1501960 (Python Code)

Max Activating Token Index: 510

Text #16

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.0908. Min Act: -0.0000

Data Index: 589402 (C4 (Web Text))

Max Activating Token Index: 423

Text #17

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.2221. Min Act: -0.0000

Data Index: 1142518 (C4 (Web Text))

Max Activating Token Index: 1013

Text #18

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.0364. Min Act: -0.0000

Data Index: 264679 (C4 (Web Text))

Max Activating Token Index: 657

Text #19

Max Range: 5.0940. Min Range: -5.0940

Max Act: 4.1904. Min Act: -0.0000

Data Index: 341375 (C4 (Web Text))

Max Activating Token Index: 537