Model: SoLU Model: 8 Layers, 4096 Neurons per Layer

Dataset: The Pile

Neuron 1694 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-8l-pile')

Text #0

Max Range: 8.4738. Min Range: -8.4738

Max Act: 8.2971. Min Act: 0.0000

Data Index: 1650497 (The Pile)

Max Activating Token Index: 266

Text #1

Max Range: 8.4738. Min Range: -8.4738

Max Act: 7.3173. Min Act: 0.0000

Data Index: 633993 (The Pile)

Max Activating Token Index: 725

Text #2

Max Range: 8.4738. Min Range: -8.4738

Max Act: 7.2407. Min Act: 0.0000

Data Index: 1918406 (The Pile)

Max Activating Token Index: 768

Text #3

Max Range: 8.4738. Min Range: -8.4738

Max Act: 7.7555. Min Act: -0.0000

Data Index: 668471 (The Pile)

Max Activating Token Index: 194

Text #4

Max Range: 8.4738. Min Range: -8.4738

Max Act: 8.4738. Min Act: -0.0000

Data Index: 150113 (The Pile)

Max Activating Token Index: 754

Text #5

Max Range: 8.4738. Min Range: -8.4738

Max Act: 7.9304. Min Act: -0.0000

Data Index: 1565417 (The Pile)

Max Activating Token Index: 459

Text #6

Max Range: 8.4738. Min Range: -8.4738

Max Act: 6.8442. Min Act: 0.0000

Data Index: 1433744 (The Pile)

Max Activating Token Index: 746

Text #7

Max Range: 8.4738. Min Range: -8.4738

Max Act: 7.2373. Min Act: -0.0000

Data Index: 1387900 (The Pile)

Max Activating Token Index: 807

Text #8

Max Range: 8.4738. Min Range: -8.4738

Max Act: 6.5173. Min Act: -0.0000

Data Index: 1336674 (The Pile)

Max Activating Token Index: 860

Text #9

Max Range: 8.4738. Min Range: -8.4738

Max Act: 6.6724. Min Act: -0.0000

Data Index: 420560 (The Pile)

Max Activating Token Index: 367

Text #10

Max Range: 8.4738. Min Range: -8.4738

Max Act: 6.9385. Min Act: -0.0000

Data Index: 923397 (The Pile)

Max Activating Token Index: 154

Text #11

Max Range: 8.4738. Min Range: -8.4738

Max Act: 6.8215. Min Act: -0.0000

Data Index: 1800145 (The Pile)

Max Activating Token Index: 1001

Text #12

Max Range: 8.4738. Min Range: -8.4738

Max Act: 6.9187. Min Act: -0.0000

Data Index: 1247167 (The Pile)

Max Activating Token Index: 251

Text #13

Max Range: 8.4738. Min Range: -8.4738

Max Act: 6.4367. Min Act: -0.0000

Data Index: 666783 (The Pile)

Max Activating Token Index: 702

Text #14

Max Range: 8.4738. Min Range: -8.4738

Max Act: 6.7284. Min Act: -0.0000

Data Index: 1402931 (The Pile)

Max Activating Token Index: 964

Text #15

Max Range: 8.4738. Min Range: -8.4738

Max Act: 6.5130. Min Act: -0.0000

Data Index: 333037 (The Pile)

Max Activating Token Index: 689

Text #16

Max Range: 8.4738. Min Range: -8.4738

Max Act: 6.6027. Min Act: -0.0000

Data Index: 1875679 (The Pile)

Max Activating Token Index: 331

Text #17

Max Range: 8.4738. Min Range: -8.4738

Max Act: 6.5831. Min Act: -0.0000

Data Index: 706379 (The Pile)

Max Activating Token Index: 433

Text #18

Max Range: 8.4738. Min Range: -8.4738

Max Act: 6.6997. Min Act: -0.0000

Data Index: 231218 (The Pile)

Max Activating Token Index: 902

Text #19

Max Range: 8.4738. Min Range: -8.4738

Max Act: 7.0896. Min Act: -0.0000

Data Index: 347073 (The Pile)

Max Activating Token Index: 932