Model: SoLU Model: 6 Layers, 3072 Neurons per Layer

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

Neuron 2704 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-6l')

Text #0

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2561. Min Act: -0.0001

Data Index: 231798 (C4 (Web Text))

Max Activating Token Index: 919

Text #1

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2406. Min Act: -0.0001

Data Index: 264419 (C4 (Web Text))

Max Activating Token Index: 131

Text #2

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2251. Min Act: -0.0001

Data Index: 1312104 (C4 (Web Text))

Max Activating Token Index: 890

Text #3

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2363. Min Act: -0.0001

Data Index: 1196954 (C4 (Web Text))

Max Activating Token Index: 539

Text #4

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2516. Min Act: -0.0001

Data Index: 452372 (C4 (Web Text))

Max Activating Token Index: 915

Text #5

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2290. Min Act: -0.0001

Data Index: 328023 (C4 (Web Text))

Max Activating Token Index: 892

Text #6

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2218. Min Act: -0.0001

Data Index: 1604766 (Python Code)

Max Activating Token Index: 112

Text #7

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2292. Min Act: -0.0001

Data Index: 861866 (C4 (Web Text))

Max Activating Token Index: 462

Text #8

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2144. Min Act: -0.0001

Data Index: 962140 (C4 (Web Text))

Max Activating Token Index: 890

Text #9

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2128. Min Act: -0.0001

Data Index: 1096072 (C4 (Web Text))

Max Activating Token Index: 642

Text #10

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2253. Min Act: -0.0001

Data Index: 703377 (C4 (Web Text))

Max Activating Token Index: 386

Text #11

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2047. Min Act: -0.0001

Data Index: 588813 (C4 (Web Text))

Max Activating Token Index: 129

Text #12

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2145. Min Act: -0.0001

Data Index: 54208 (C4 (Web Text))

Max Activating Token Index: 920

Text #13

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.1948. Min Act: -0.0001

Data Index: 986023 (C4 (Web Text))

Max Activating Token Index: 706

Text #14

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.1998. Min Act: -0.0001

Data Index: 1333506 (C4 (Web Text))

Max Activating Token Index: 597

Text #15

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.1920. Min Act: -0.0001

Data Index: 835417 (C4 (Web Text))

Max Activating Token Index: 32

Text #16

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2017. Min Act: -0.0001

Data Index: 1466386 (Python Code)

Max Activating Token Index: 832

Text #17

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2084. Min Act: -0.0001

Data Index: 185427 (C4 (Web Text))

Max Activating Token Index: 380

Text #18

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.1954. Min Act: -0.0001

Data Index: 768652 (C4 (Web Text))

Max Activating Token Index: 173

Text #19

Max Range: 0.2561. Min Range: -0.2561

Max Act: 0.2022. Min Act: -0.0001

Data Index: 923895 (C4 (Web Text))

Max Activating Token Index: 884