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

Dataset: The Pile

Neuron 728 in Layer 7

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 7.8959. Min Range: -7.8959

Max Act: 7.8959. Min Act: -0.0000

Data Index: 1709066 (The Pile)

Max Activating Token Index: 160

Text #1

Max Range: 7.8959. Min Range: -7.8959

Max Act: 7.4438. Min Act: -0.0000

Data Index: 508423 (The Pile)

Max Activating Token Index: 684

Text #2

Max Range: 7.8959. Min Range: -7.8959

Max Act: 7.3961. Min Act: -0.0000

Data Index: 428444 (The Pile)

Max Activating Token Index: 238

Text #3

Max Range: 7.8959. Min Range: -7.8959

Max Act: 7.6315. Min Act: -0.0000

Data Index: 1383044 (The Pile)

Max Activating Token Index: 569

Text #4

Max Range: 7.8959. Min Range: -7.8959

Max Act: 7.3699. Min Act: -0.0000

Data Index: 1525489 (The Pile)

Max Activating Token Index: 203

Text #5

Max Range: 7.8959. Min Range: -7.8959

Max Act: 7.1899. Min Act: -0.0000

Data Index: 13228 (The Pile)

Max Activating Token Index: 519

Text #6

Max Range: 7.8959. Min Range: -7.8959

Max Act: 6.8001. Min Act: -0.0000

Data Index: 1665419 (The Pile)

Max Activating Token Index: 228

Text #7

Max Range: 7.8959. Min Range: -7.8959

Max Act: 6.7436. Min Act: -0.0000

Data Index: 749472 (The Pile)

Max Activating Token Index: 1000

Text #8

Max Range: 7.8959. Min Range: -7.8959

Max Act: 6.9180. Min Act: -0.0000

Data Index: 399339 (The Pile)

Max Activating Token Index: 313

Text #9

Max Range: 7.8959. Min Range: -7.8959

Max Act: 7.0181. Min Act: -0.0000

Data Index: 1598810 (The Pile)

Max Activating Token Index: 293

Text #10

Max Range: 7.8959. Min Range: -7.8959

Max Act: 7.0169. Min Act: -0.0000

Data Index: 631197 (The Pile)

Max Activating Token Index: 184

Text #11

Max Range: 7.8959. Min Range: -7.8959

Max Act: 6.4368. Min Act: -0.0000

Data Index: 1722801 (The Pile)

Max Activating Token Index: 109

Text #12

Max Range: 7.8959. Min Range: -7.8959

Max Act: 7.2978. Min Act: -0.0000

Data Index: 1072879 (The Pile)

Max Activating Token Index: 671

Text #13

Max Range: 7.8959. Min Range: -7.8959

Max Act: 7.1305. Min Act: -0.0000

Data Index: 1697099 (The Pile)

Max Activating Token Index: 938

Text #14

Max Range: 7.8959. Min Range: -7.8959

Max Act: 7.1484. Min Act: -0.0000

Data Index: 1653245 (The Pile)

Max Activating Token Index: 58

Text #15

Max Range: 7.8959. Min Range: -7.8959

Max Act: 6.3433. Min Act: -0.0000

Data Index: 144041 (The Pile)

Max Activating Token Index: 840

Text #16

Max Range: 7.8959. Min Range: -7.8959

Max Act: 6.3975. Min Act: -0.0000

Data Index: 831090 (The Pile)

Max Activating Token Index: 322

Text #17

Max Range: 7.8959. Min Range: -7.8959

Max Act: 6.7713. Min Act: -0.0000

Data Index: 1773421 (The Pile)

Max Activating Token Index: 516

Text #18

Max Range: 7.8959. Min Range: -7.8959

Max Act: 6.5772. Min Act: -0.0000

Data Index: 1817502 (The Pile)

Max Activating Token Index: 166

Text #19

Max Range: 7.8959. Min Range: -7.8959

Max Act: 6.3313. Min Act: -0.0000

Data Index: 199758 (The Pile)

Max Activating Token Index: 431