Model: SoLU Model: 4 Layers, 2048 Neurons per Layer

Dataset: 80% C4 (Web Text) and 20% Python Code

Neuron 654 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('solu-4l')

Text #0

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0189. Min Act: -0.0006

Data Index: 774474 (C4 (Web Text))

Max Activating Token Index: 1002

Text #1

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0177. Min Act: -0.0006

Data Index: 1343524 (C4 (Web Text))

Max Activating Token Index: 135

Text #2

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0162. Min Act: -0.0006

Data Index: 97131 (C4 (Web Text))

Max Activating Token Index: 1022

Text #3

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0144. Min Act: -0.0006

Data Index: 858005 (C4 (Web Text))

Max Activating Token Index: 960

Text #4

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0141. Min Act: -0.0006

Data Index: 25852 (C4 (Web Text))

Max Activating Token Index: 532

Text #5

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0144. Min Act: -0.0006

Data Index: 83587 (C4 (Web Text))

Max Activating Token Index: 943

Text #6

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0149. Min Act: -0.0006

Data Index: 1207715 (C4 (Web Text))

Max Activating Token Index: 1003

Text #7

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0135. Min Act: -0.0005

Data Index: 1363758 (Python Code)

Max Activating Token Index: 1023

Text #8

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0140. Min Act: -0.0006

Data Index: 768220 (C4 (Web Text))

Max Activating Token Index: 1006

Text #9

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0142. Min Act: -0.0006

Data Index: 556052 (C4 (Web Text))

Max Activating Token Index: 986

Text #10

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0140. Min Act: -0.0005

Data Index: 1555720 (Python Code)

Max Activating Token Index: 1023

Text #11

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0141. Min Act: -0.0006

Data Index: 1011144 (C4 (Web Text))

Max Activating Token Index: 588

Text #12

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0132. Min Act: -0.0006

Data Index: 131848 (C4 (Web Text))

Max Activating Token Index: 1015

Text #13

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0142. Min Act: -0.0006

Data Index: 1545619 (Python Code)

Max Activating Token Index: 752

Text #14

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0124. Min Act: -0.0006

Data Index: 206754 (C4 (Web Text))

Max Activating Token Index: 988

Text #15

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0139. Min Act: -0.0006

Data Index: 383282 (C4 (Web Text))

Max Activating Token Index: 763

Text #16

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0123. Min Act: -0.0006

Data Index: 824809 (C4 (Web Text))

Max Activating Token Index: 635

Text #17

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0134. Min Act: -0.0006

Data Index: 601393 (C4 (Web Text))

Max Activating Token Index: 1015

Text #18

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0131. Min Act: -0.0006

Data Index: 334024 (C4 (Web Text))

Max Activating Token Index: 845

Text #19

Max Range: 0.0189. Min Range: -0.0189

Max Act: 0.0125. Min Act: -0.0006

Data Index: 1077961 (C4 (Web Text))

Max Activating Token Index: 727