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

Dataset: The Pile

Neuron 4098 in Layer 6

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-10l-pile')

Text #0

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.7976. Min Act: -0.0000

Data Index: 191239 (The Pile)

Max Activating Token Index: 146

Text #1

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.6293. Min Act: -0.0000

Data Index: 1266187 (The Pile)

Max Activating Token Index: 952

Text #2

Max Range: 9.3117. Min Range: -9.3117

Max Act: 9.3117. Min Act: -0.0000

Data Index: 1530100 (The Pile)

Max Activating Token Index: 381

Text #3

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.9248. Min Act: -0.0000

Data Index: 1854204 (The Pile)

Max Activating Token Index: 909

Text #4

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.7810. Min Act: -0.0000

Data Index: 214809 (The Pile)

Max Activating Token Index: 163

Text #5

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.4792. Min Act: -0.0000

Data Index: 294575 (The Pile)

Max Activating Token Index: 148

Text #6

Max Range: 9.3117. Min Range: -9.3117

Max Act: 9.0500. Min Act: -0.0000

Data Index: 281551 (The Pile)

Max Activating Token Index: 23

Text #7

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.3636. Min Act: -0.0000

Data Index: 1147947 (The Pile)

Max Activating Token Index: 236

Text #8

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.8456. Min Act: -0.0000

Data Index: 225658 (The Pile)

Max Activating Token Index: 292

Text #9

Max Range: 9.3117. Min Range: -9.3117

Max Act: 7.6433. Min Act: -0.0000

Data Index: 947103 (The Pile)

Max Activating Token Index: 698

Text #10

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.2271. Min Act: -0.0000

Data Index: 930174 (The Pile)

Max Activating Token Index: 2

Text #11

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.2271. Min Act: -0.0000

Data Index: 528900 (The Pile)

Max Activating Token Index: 2

Text #12

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.1983. Min Act: -0.0000

Data Index: 1052695 (The Pile)

Max Activating Token Index: 791

Text #13

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.8567. Min Act: -0.0000

Data Index: 1718062 (The Pile)

Max Activating Token Index: 61

Text #14

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.6712. Min Act: -0.0000

Data Index: 1561201 (The Pile)

Max Activating Token Index: 271

Text #15

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.4830. Min Act: -0.0000

Data Index: 871178 (The Pile)

Max Activating Token Index: 120

Text #16

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.6512. Min Act: -0.0000

Data Index: 1358301 (The Pile)

Max Activating Token Index: 426

Text #17

Max Range: 9.3117. Min Range: -9.3117

Max Act: 8.3056. Min Act: -0.0000

Data Index: 591698 (The Pile)

Max Activating Token Index: 854

Text #18

Max Range: 9.3117. Min Range: -9.3117

Max Act: 7.9227. Min Act: -0.0000

Data Index: 317315 (The Pile)

Max Activating Token Index: 174

Text #19

Max Range: 9.3117. Min Range: -9.3117

Max Act: 7.9449. Min Act: -0.0000

Data Index: 417631 (The Pile)

Max Activating Token Index: 902