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

Dataset: The Pile

Neuron 553 in Layer 5

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.9784. Min Act: -0.0000

Data Index: 1140887 (The Pile)

Max Activating Token Index: 431

Text #1

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.6891. Min Act: -0.0000

Data Index: 1145276 (The Pile)

Max Activating Token Index: 869

Text #2

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.6880. Min Act: -0.0000

Data Index: 189801 (The Pile)

Max Activating Token Index: 116

Text #3

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.0996. Min Act: -0.0000

Data Index: 1928833 (The Pile)

Max Activating Token Index: 62

Text #4

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.3910. Min Act: -0.0000

Data Index: 1211214 (The Pile)

Max Activating Token Index: 139

Text #5

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.3795. Min Act: -0.0000

Data Index: 1411536 (The Pile)

Max Activating Token Index: 967

Text #6

Max Range: 3.9784. Min Range: -3.9784

Max Act: 2.9002. Min Act: -0.0000

Data Index: 324409 (The Pile)

Max Activating Token Index: 768

Text #7

Max Range: 3.9784. Min Range: -3.9784

Max Act: 2.8934. Min Act: -0.0000

Data Index: 1948204 (The Pile)

Max Activating Token Index: 563

Text #8

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.1381. Min Act: -0.0000

Data Index: 1026044 (The Pile)

Max Activating Token Index: 878

Text #9

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.1103. Min Act: -0.0000

Data Index: 1525012 (The Pile)

Max Activating Token Index: 408

Text #10

Max Range: 3.9784. Min Range: -3.9784

Max Act: 2.9468. Min Act: -0.0000

Data Index: 324288 (The Pile)

Max Activating Token Index: 196

Text #11

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.3669. Min Act: -0.0000

Data Index: 98720 (The Pile)

Max Activating Token Index: 510

Text #12

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.3891. Min Act: -0.0000

Data Index: 123528 (The Pile)

Max Activating Token Index: 450

Text #13

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.0656. Min Act: -0.0000

Data Index: 1320043 (The Pile)

Max Activating Token Index: 460

Text #14

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.0393. Min Act: -0.0000

Data Index: 300652 (The Pile)

Max Activating Token Index: 92

Text #15

Max Range: 3.9784. Min Range: -3.9784

Max Act: 2.6610. Min Act: -0.0000

Data Index: 867774 (The Pile)

Max Activating Token Index: 382

Text #16

Max Range: 3.9784. Min Range: -3.9784

Max Act: 2.8885. Min Act: -0.0000

Data Index: 715534 (The Pile)

Max Activating Token Index: 774

Text #17

Max Range: 3.9784. Min Range: -3.9784

Max Act: 2.8782. Min Act: -0.0000

Data Index: 717452 (The Pile)

Max Activating Token Index: 764

Text #18

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.0109. Min Act: -0.0000

Data Index: 986859 (The Pile)

Max Activating Token Index: 664

Text #19

Max Range: 3.9784. Min Range: -3.9784

Max Act: 3.2780. Min Act: -0.0000

Data Index: 367863 (The Pile)

Max Activating Token Index: 688