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

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

Neuron 3271 in Layer 9

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.4286. Min Act: -0.0000

Data Index: 532569 (C4 (Web Text))

Max Activating Token Index: 65

Text #1

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.4767. Min Act: -0.0000

Data Index: 162030 (C4 (Web Text))

Max Activating Token Index: 618

Text #2

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.4056. Min Act: -0.0000

Data Index: 1360110 (Python Code)

Max Activating Token Index: 709

Text #3

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.3622. Min Act: -0.0000

Data Index: 1457359 (Python Code)

Max Activating Token Index: 628

Text #4

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.3692. Min Act: -0.0000

Data Index: 564925 (C4 (Web Text))

Max Activating Token Index: 210

Text #5

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.3480. Min Act: -0.0000

Data Index: 268283 (C4 (Web Text))

Max Activating Token Index: 38

Text #6

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.3400. Min Act: -0.0000

Data Index: 1569773 (Python Code)

Max Activating Token Index: 384

Text #7

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.3502. Min Act: -0.0000

Data Index: 444521 (C4 (Web Text))

Max Activating Token Index: 123

Text #8

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.3001. Min Act: -0.0000

Data Index: 1613123 (Python Code)

Max Activating Token Index: 16

Text #9

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.2931. Min Act: -0.0000

Data Index: 310816 (C4 (Web Text))

Max Activating Token Index: 370

Text #10

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.3135. Min Act: -0.0000

Data Index: 1493168 (Python Code)

Max Activating Token Index: 255

Text #11

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.2748. Min Act: -0.0000

Data Index: 1167313 (C4 (Web Text))

Max Activating Token Index: 252

Text #12

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.2886. Min Act: -0.0000

Data Index: 1194113 (C4 (Web Text))

Max Activating Token Index: 326

Text #13

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.2809. Min Act: -0.0000

Data Index: 779287 (C4 (Web Text))

Max Activating Token Index: 339

Text #14

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.2839. Min Act: -0.0000

Data Index: 813294 (C4 (Web Text))

Max Activating Token Index: 241

Text #15

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.2860. Min Act: -0.0000

Data Index: 778589 (C4 (Web Text))

Max Activating Token Index: 380

Text #16

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.2641. Min Act: -0.0000

Data Index: 1611195 (Python Code)

Max Activating Token Index: 124

Text #17

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.2821. Min Act: -0.0000

Data Index: 21131 (C4 (Web Text))

Max Activating Token Index: 304

Text #18

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.2646. Min Act: -0.0000

Data Index: 909333 (C4 (Web Text))

Max Activating Token Index: 449

Text #19

Max Range: 0.4767. Min Range: -0.4767

Max Act: 0.3019. Min Act: -0.0000

Data Index: 384120 (C4 (Web Text))

Max Activating Token Index: 299