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

Dataset: The Pile

Neuron 9 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 5.1745. Min Range: -5.1745

Max Act: 4.8847. Min Act: -0.0000

Data Index: 1994852 (The Pile)

Max Activating Token Index: 343

Text #1

Max Range: 5.1745. Min Range: -5.1745

Max Act: 4.9088. Min Act: 0.0000

Data Index: 733347 (The Pile)

Max Activating Token Index: 791

Text #2

Max Range: 5.1745. Min Range: -5.1745

Max Act: 5.1745. Min Act: -0.0000

Data Index: 734245 (The Pile)

Max Activating Token Index: 128

Text #3

Max Range: 5.1745. Min Range: -5.1745

Max Act: 4.8329. Min Act: -0.0000

Data Index: 752129 (The Pile)

Max Activating Token Index: 721

Text #4

Max Range: 5.1745. Min Range: -5.1745

Max Act: 4.4685. Min Act: -0.0000

Data Index: 470408 (The Pile)

Max Activating Token Index: 994

Text #5

Max Range: 5.1745. Min Range: -5.1745

Max Act: 4.0836. Min Act: -0.0000

Data Index: 1726421 (The Pile)

Max Activating Token Index: 715

Text #6

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.6146. Min Act: -0.0000

Data Index: 352008 (The Pile)

Max Activating Token Index: 485

Text #7

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.8927. Min Act: -0.0000

Data Index: 1846499 (The Pile)

Max Activating Token Index: 666

Text #8

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.1444. Min Act: -0.0000

Data Index: 554064 (The Pile)

Max Activating Token Index: 452

Text #9

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.9238. Min Act: -0.0000

Data Index: 1697097 (The Pile)

Max Activating Token Index: 188

Text #10

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.5502. Min Act: -0.0000

Data Index: 620927 (The Pile)

Max Activating Token Index: 403

Text #11

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.5304. Min Act: -0.0000

Data Index: 1802246 (The Pile)

Max Activating Token Index: 681

Text #12

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.6884. Min Act: -0.0000

Data Index: 1033780 (The Pile)

Max Activating Token Index: 521

Text #13

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.3351. Min Act: -0.0000

Data Index: 765595 (The Pile)

Max Activating Token Index: 254

Text #14

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.0958. Min Act: -0.0000

Data Index: 398823 (The Pile)

Max Activating Token Index: 1020

Text #15

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.4591. Min Act: -0.0000

Data Index: 799842 (The Pile)

Max Activating Token Index: 447

Text #16

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.3442. Min Act: -0.0000

Data Index: 903571 (The Pile)

Max Activating Token Index: 842

Text #17

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.4958. Min Act: -0.0000

Data Index: 91056 (The Pile)

Max Activating Token Index: 550

Text #18

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.3185. Min Act: -0.0000

Data Index: 1565776 (The Pile)

Max Activating Token Index: 662

Text #19

Max Range: 5.1745. Min Range: -5.1745

Max Act: 3.6218. Min Act: -0.0000

Data Index: 1776124 (The Pile)

Max Activating Token Index: 549