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

Dataset: The Pile

Neuron 4703 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.5145. Min Range: -5.5145

Max Act: 5.5145. Min Act: -0.0000

Data Index: 1326521 (The Pile)

Max Activating Token Index: 798

Text #1

Max Range: 5.5145. Min Range: -5.5145

Max Act: 5.2824. Min Act: -0.0000

Data Index: 83454 (The Pile)

Max Activating Token Index: 396

Text #2

Max Range: 5.5145. Min Range: -5.5145

Max Act: 5.2492. Min Act: -0.0000

Data Index: 643897 (The Pile)

Max Activating Token Index: 10

Text #3

Max Range: 5.5145. Min Range: -5.5145

Max Act: 5.1205. Min Act: -0.0000

Data Index: 1221419 (The Pile)

Max Activating Token Index: 856

Text #4

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.8488. Min Act: -0.0000

Data Index: 943937 (The Pile)

Max Activating Token Index: 379

Text #5

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.6262. Min Act: -0.0000

Data Index: 723903 (The Pile)

Max Activating Token Index: 894

Text #6

Max Range: 5.5145. Min Range: -5.5145

Max Act: 5.0088. Min Act: -0.0000

Data Index: 1132625 (The Pile)

Max Activating Token Index: 256

Text #7

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.7914. Min Act: -0.0000

Data Index: 106995 (The Pile)

Max Activating Token Index: 19

Text #8

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.7597. Min Act: -0.0000

Data Index: 589006 (The Pile)

Max Activating Token Index: 268

Text #9

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.5997. Min Act: -0.0000

Data Index: 642023 (The Pile)

Max Activating Token Index: 863

Text #10

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.5374. Min Act: -0.0000

Data Index: 186450 (The Pile)

Max Activating Token Index: 313

Text #11

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.6848. Min Act: -0.0000

Data Index: 1848961 (The Pile)

Max Activating Token Index: 7

Text #12

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.4919. Min Act: -0.0000

Data Index: 774397 (The Pile)

Max Activating Token Index: 1008

Text #13

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.4216. Min Act: -0.0000

Data Index: 1239547 (The Pile)

Max Activating Token Index: 327

Text #14

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.4114. Min Act: -0.0000

Data Index: 1106948 (The Pile)

Max Activating Token Index: 1008

Text #15

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.4773. Min Act: -0.0000

Data Index: 754425 (The Pile)

Max Activating Token Index: 689

Text #16

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.4920. Min Act: -0.0000

Data Index: 1895541 (The Pile)

Max Activating Token Index: 399

Text #17

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.7591. Min Act: -0.0000

Data Index: 630745 (The Pile)

Max Activating Token Index: 687

Text #18

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.5548. Min Act: -0.0000

Data Index: 265966 (The Pile)

Max Activating Token Index: 639

Text #19

Max Range: 5.5145. Min Range: -5.5145

Max Act: 4.4328. Min Act: -0.0000

Data Index: 1820538 (The Pile)

Max Activating Token Index: 914