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

Dataset: The Pile

Neuron 4293 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: 9.9777. Min Range: -9.9777

Max Act: 9.9777. Min Act: -0.0000

Data Index: 1907212 (The Pile)

Max Activating Token Index: 94

Text #1

Max Range: 9.9777. Min Range: -9.9777

Max Act: 9.3942. Min Act: -0.0000

Data Index: 1761530 (The Pile)

Max Activating Token Index: 345

Text #2

Max Range: 9.9777. Min Range: -9.9777

Max Act: 9.5573. Min Act: -0.0000

Data Index: 955882 (The Pile)

Max Activating Token Index: 412

Text #3

Max Range: 9.9777. Min Range: -9.9777

Max Act: 9.0719. Min Act: -0.0000

Data Index: 1453452 (The Pile)

Max Activating Token Index: 94

Text #4

Max Range: 9.9777. Min Range: -9.9777

Max Act: 9.3093. Min Act: -0.0000

Data Index: 1447676 (The Pile)

Max Activating Token Index: 99

Text #5

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.9185. Min Act: -0.0000

Data Index: 75265 (The Pile)

Max Activating Token Index: 108

Text #6

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.0275. Min Act: -0.0000

Data Index: 1227558 (The Pile)

Max Activating Token Index: 56

Text #7

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.0418. Min Act: -0.0000

Data Index: 675865 (The Pile)

Max Activating Token Index: 454

Text #8

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.3425. Min Act: -0.0000

Data Index: 231684 (The Pile)

Max Activating Token Index: 308

Text #9

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.7205. Min Act: -0.0000

Data Index: 780129 (The Pile)

Max Activating Token Index: 118

Text #10

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.7494. Min Act: -0.0000

Data Index: 1580592 (The Pile)

Max Activating Token Index: 352

Text #11

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.5904. Min Act: -0.0000

Data Index: 1915944 (The Pile)

Max Activating Token Index: 194

Text #12

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.1535. Min Act: -0.0000

Data Index: 26753 (The Pile)

Max Activating Token Index: 637

Text #13

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.7676. Min Act: -0.0000

Data Index: 1481807 (The Pile)

Max Activating Token Index: 327

Text #14

Max Range: 9.9777. Min Range: -9.9777

Max Act: 7.8402. Min Act: -0.0000

Data Index: 1389119 (The Pile)

Max Activating Token Index: 251

Text #15

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.4830. Min Act: -0.0000

Data Index: 1352025 (The Pile)

Max Activating Token Index: 483

Text #16

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.5807. Min Act: -0.0000

Data Index: 1578779 (The Pile)

Max Activating Token Index: 574

Text #17

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.2057. Min Act: -0.0000

Data Index: 1060910 (The Pile)

Max Activating Token Index: 586

Text #18

Max Range: 9.9777. Min Range: -9.9777

Max Act: 7.8624. Min Act: -0.0000

Data Index: 510519 (The Pile)

Max Activating Token Index: 124

Text #19

Max Range: 9.9777. Min Range: -9.9777

Max Act: 8.3090. Min Act: -0.0000

Data Index: 884893 (The Pile)

Max Activating Token Index: 246