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

Dataset: The Pile

Neuron 3426 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: 9.3951. Min Range: -9.3951

Max Act: 9.2469. Min Act: -0.0000

Data Index: 1180869 (The Pile)

Max Activating Token Index: 129

Text #1

Max Range: 9.3951. Min Range: -9.3951

Max Act: 9.2346. Min Act: -0.0000

Data Index: 857104 (The Pile)

Max Activating Token Index: 105

Text #2

Max Range: 9.3951. Min Range: -9.3951

Max Act: 9.1335. Min Act: -0.0001

Data Index: 1191269 (The Pile)

Max Activating Token Index: 158

Text #3

Max Range: 9.3951. Min Range: -9.3951

Max Act: 9.3951. Min Act: -0.0000

Data Index: 1234953 (The Pile)

Max Activating Token Index: 117

Text #4

Max Range: 9.3951. Min Range: -9.3951

Max Act: 8.8746. Min Act: -0.0000

Data Index: 1592775 (The Pile)

Max Activating Token Index: 172

Text #5

Max Range: 9.3951. Min Range: -9.3951

Max Act: 9.0490. Min Act: -0.0000

Data Index: 1151500 (The Pile)

Max Activating Token Index: 175

Text #6

Max Range: 9.3951. Min Range: -9.3951

Max Act: 9.0916. Min Act: -0.0000

Data Index: 1811281 (The Pile)

Max Activating Token Index: 208

Text #7

Max Range: 9.3951. Min Range: -9.3951

Max Act: 8.4576. Min Act: -0.0000

Data Index: 137163 (The Pile)

Max Activating Token Index: 496

Text #8

Max Range: 9.3951. Min Range: -9.3951

Max Act: 8.5160. Min Act: -0.0000

Data Index: 772892 (The Pile)

Max Activating Token Index: 480

Text #9

Max Range: 9.3951. Min Range: -9.3951

Max Act: 9.0967. Min Act: -0.0000

Data Index: 1295703 (The Pile)

Max Activating Token Index: 156

Text #10

Max Range: 9.3951. Min Range: -9.3951

Max Act: 9.1499. Min Act: -0.0000

Data Index: 422411 (The Pile)

Max Activating Token Index: 151

Text #11

Max Range: 9.3951. Min Range: -9.3951

Max Act: 8.5715. Min Act: -0.0000

Data Index: 525098 (The Pile)

Max Activating Token Index: 191

Text #12

Max Range: 9.3951. Min Range: -9.3951

Max Act: 8.8679. Min Act: -0.0000

Data Index: 315881 (The Pile)

Max Activating Token Index: 568

Text #13

Max Range: 9.3951. Min Range: -9.3951

Max Act: 8.4273. Min Act: -0.0000

Data Index: 138236 (The Pile)

Max Activating Token Index: 680

Text #14

Max Range: 9.3951. Min Range: -9.3951

Max Act: 8.8684. Min Act: -0.0001

Data Index: 869840 (The Pile)

Max Activating Token Index: 172

Text #15

Max Range: 9.3951. Min Range: -9.3951

Max Act: 8.6632. Min Act: -0.0000

Data Index: 57532 (The Pile)

Max Activating Token Index: 193

Text #16

Max Range: 9.3951. Min Range: -9.3951

Max Act: 8.8154. Min Act: -0.0000

Data Index: 445477 (The Pile)

Max Activating Token Index: 187

Text #17

Max Range: 9.3951. Min Range: -9.3951

Max Act: 8.1250. Min Act: -0.0000

Data Index: 1880904 (The Pile)

Max Activating Token Index: 602

Text #18

Max Range: 9.3951. Min Range: -9.3951

Max Act: 8.1776. Min Act: -0.0000

Data Index: 231468 (The Pile)

Max Activating Token Index: 78

Text #19

Max Range: 9.3951. Min Range: -9.3951

Max Act: 8.3514. Min Act: -0.0000

Data Index: 1725527 (The Pile)

Max Activating Token Index: 544