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

Dataset: The Pile

Neuron 471 in Layer 8

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 8.3095. Min Range: -8.3095

Max Act: 8.3095. Min Act: -0.0000

Data Index: 1679999 (The Pile)

Max Activating Token Index: 824

Text #1

Max Range: 8.3095. Min Range: -8.3095

Max Act: 8.1305. Min Act: -0.0000

Data Index: 1363385 (The Pile)

Max Activating Token Index: 856

Text #2

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.9522. Min Act: 0.0004

Data Index: 1210941 (The Pile)

Max Activating Token Index: 732

Text #3

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.8062. Min Act: 0.0008

Data Index: 1734701 (The Pile)

Max Activating Token Index: 606

Text #4

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.6848. Min Act: -0.0000

Data Index: 362943 (The Pile)

Max Activating Token Index: 669

Text #5

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.4666. Min Act: -0.0000

Data Index: 671985 (The Pile)

Max Activating Token Index: 207

Text #6

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.4393. Min Act: 0.0002

Data Index: 1608378 (The Pile)

Max Activating Token Index: 760

Text #7

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.4336. Min Act: -0.0000

Data Index: 538912 (The Pile)

Max Activating Token Index: 984

Text #8

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.5319. Min Act: -0.0000

Data Index: 1803138 (The Pile)

Max Activating Token Index: 168

Text #9

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.4295. Min Act: -0.0000

Data Index: 1872847 (The Pile)

Max Activating Token Index: 459

Text #10

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.3939. Min Act: -0.0000

Data Index: 1731154 (The Pile)

Max Activating Token Index: 191

Text #11

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.6109. Min Act: -0.0000

Data Index: 787206 (The Pile)

Max Activating Token Index: 700

Text #12

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.5987. Min Act: -0.0000

Data Index: 604783 (The Pile)

Max Activating Token Index: 272

Text #13

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.0903. Min Act: -0.0000

Data Index: 1710354 (The Pile)

Max Activating Token Index: 760

Text #14

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.4539. Min Act: -0.0000

Data Index: 1600488 (The Pile)

Max Activating Token Index: 973

Text #15

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.4131. Min Act: -0.0000

Data Index: 1943868 (The Pile)

Max Activating Token Index: 640

Text #16

Max Range: 8.3095. Min Range: -8.3095

Max Act: 6.9873. Min Act: -0.0000

Data Index: 324125 (The Pile)

Max Activating Token Index: 873

Text #17

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.1756. Min Act: 0.0008

Data Index: 417570 (The Pile)

Max Activating Token Index: 52

Text #18

Max Range: 8.3095. Min Range: -8.3095

Max Act: 6.9165. Min Act: -0.0000

Data Index: 1390424 (The Pile)

Max Activating Token Index: 776

Text #19

Max Range: 8.3095. Min Range: -8.3095

Max Act: 7.1119. Min Act: -0.0000

Data Index: 1741855 (The Pile)

Max Activating Token Index: 358