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

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

Neuron 3065 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.2298. Min Act: -0.0003

Data Index: 101732 (C4 (Web Text))

Max Activating Token Index: 700

Text #1

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.2253. Min Act: -0.0003

Data Index: 1093018 (C4 (Web Text))

Max Activating Token Index: 371

Text #2

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.2129. Min Act: -0.0003

Data Index: 1354943 (C4 (Web Text))

Max Activating Token Index: 905

Text #3

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.1908. Min Act: -0.0003

Data Index: 1150197 (C4 (Web Text))

Max Activating Token Index: 106

Text #4

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.2070. Min Act: -0.0003

Data Index: 644089 (C4 (Web Text))

Max Activating Token Index: 828

Text #5

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.2062. Min Act: -0.0003

Data Index: 544303 (C4 (Web Text))

Max Activating Token Index: 303

Text #6

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.1960. Min Act: -0.0003

Data Index: 636267 (C4 (Web Text))

Max Activating Token Index: 831

Text #7

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.1978. Min Act: -0.0003

Data Index: 348202 (C4 (Web Text))

Max Activating Token Index: 207

Text #8

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.2115. Min Act: -0.0003

Data Index: 717842 (C4 (Web Text))

Max Activating Token Index: 232

Text #9

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.2115. Min Act: -0.0003

Data Index: 618047 (C4 (Web Text))

Max Activating Token Index: 624

Text #10

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.1944. Min Act: -0.0003

Data Index: 1631693 (Python Code)

Max Activating Token Index: 511

Text #11

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.2049. Min Act: -0.0003

Data Index: 721417 (C4 (Web Text))

Max Activating Token Index: 160

Text #12

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.1960. Min Act: -0.0003

Data Index: 938042 (C4 (Web Text))

Max Activating Token Index: 182

Text #13

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.1902. Min Act: -0.0003

Data Index: 296796 (C4 (Web Text))

Max Activating Token Index: 689

Text #14

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.2032. Min Act: -0.0003

Data Index: 357655 (C4 (Web Text))

Max Activating Token Index: 290

Text #15

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.1807. Min Act: -0.0003

Data Index: 548454 (C4 (Web Text))

Max Activating Token Index: 473

Text #16

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.1743. Min Act: -0.0003

Data Index: 129890 (C4 (Web Text))

Max Activating Token Index: 571

Text #17

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.1852. Min Act: -0.0003

Data Index: 686187 (C4 (Web Text))

Max Activating Token Index: 15

Text #18

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.1754. Min Act: -0.0003

Data Index: 1618793 (Python Code)

Max Activating Token Index: 328

Text #19

Max Range: 0.2298. Min Range: -0.2298

Max Act: 0.1849. Min Act: -0.0003

Data Index: 1222314 (C4 (Web Text))

Max Activating Token Index: 487