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

Dataset: The Pile

Neuron 4608 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: 4.3945. Min Range: -4.3945

Max Act: 4.3945. Min Act: -0.0000

Data Index: 1607027 (The Pile)

Max Activating Token Index: 8

Text #1

Max Range: 4.3945. Min Range: -4.3945

Max Act: 4.1694. Min Act: -0.0000

Data Index: 592891 (The Pile)

Max Activating Token Index: 454

Text #2

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.9900. Min Act: -0.0000

Data Index: 1817652 (The Pile)

Max Activating Token Index: 816

Text #3

Max Range: 4.3945. Min Range: -4.3945

Max Act: 4.0907. Min Act: -0.0000

Data Index: 185883 (The Pile)

Max Activating Token Index: 60

Text #4

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.5727. Min Act: -0.0000

Data Index: 526239 (The Pile)

Max Activating Token Index: 50

Text #5

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.9975. Min Act: -0.0000

Data Index: 1077063 (The Pile)

Max Activating Token Index: 184

Text #6

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.4942. Min Act: -0.0000

Data Index: 1674912 (The Pile)

Max Activating Token Index: 203

Text #7

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.5842. Min Act: -0.0000

Data Index: 1429301 (The Pile)

Max Activating Token Index: 723

Text #8

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.4890. Min Act: -0.0000

Data Index: 888939 (The Pile)

Max Activating Token Index: 98

Text #9

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.4467. Min Act: -0.0000

Data Index: 1992394 (The Pile)

Max Activating Token Index: 146

Text #10

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.3855. Min Act: -0.0000

Data Index: 1198285 (The Pile)

Max Activating Token Index: 835

Text #11

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.3764. Min Act: -0.0000

Data Index: 473612 (The Pile)

Max Activating Token Index: 835

Text #12

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.4891. Min Act: -0.0000

Data Index: 1814482 (The Pile)

Max Activating Token Index: 277

Text #13

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.3086. Min Act: -0.0000

Data Index: 849908 (The Pile)

Max Activating Token Index: 541

Text #14

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.0731. Min Act: -0.0000

Data Index: 74781 (The Pile)

Max Activating Token Index: 853

Text #15

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.1735. Min Act: -0.0000

Data Index: 1181287 (The Pile)

Max Activating Token Index: 1023

Text #16

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.3231. Min Act: -0.0000

Data Index: 791913 (The Pile)

Max Activating Token Index: 556

Text #17

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.4989. Min Act: -0.0000

Data Index: 606231 (The Pile)

Max Activating Token Index: 816

Text #18

Max Range: 4.3945. Min Range: -4.3945

Max Act: 3.0065. Min Act: -0.0000

Data Index: 1115333 (The Pile)

Max Activating Token Index: 263

Text #19

Max Range: 4.3945. Min Range: -4.3945

Max Act: 2.9826. Min Act: -0.0000

Data Index: 704455 (The Pile)

Max Activating Token Index: 422