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

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

Neuron 3027 in Layer 4

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.1170. Min Act: -0.0001

Data Index: 292121 (C4 (Web Text))

Max Activating Token Index: 10

Text #1

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.1076. Min Act: -0.0001

Data Index: 493911 (C4 (Web Text))

Max Activating Token Index: 401

Text #2

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0956. Min Act: -0.0001

Data Index: 1352557 (C4 (Web Text))

Max Activating Token Index: 10

Text #3

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0957. Min Act: -0.0001

Data Index: 1638996 (Python Code)

Max Activating Token Index: 695

Text #4

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0874. Min Act: -0.0001

Data Index: 993475 (C4 (Web Text))

Max Activating Token Index: 755

Text #5

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0921. Min Act: -0.0001

Data Index: 1028996 (C4 (Web Text))

Max Activating Token Index: 445

Text #6

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0965. Min Act: -0.0001

Data Index: 1381613 (Python Code)

Max Activating Token Index: 637

Text #7

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0953. Min Act: -0.0001

Data Index: 1377029 (Python Code)

Max Activating Token Index: 659

Text #8

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0905. Min Act: -0.0001

Data Index: 1442771 (Python Code)

Max Activating Token Index: 848

Text #9

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0851. Min Act: -0.0001

Data Index: 998956 (C4 (Web Text))

Max Activating Token Index: 669

Text #10

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0957. Min Act: -0.0001

Data Index: 1360278 (Python Code)

Max Activating Token Index: 139

Text #11

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0962. Min Act: -0.0001

Data Index: 1482419 (Python Code)

Max Activating Token Index: 209

Text #12

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0855. Min Act: -0.0001

Data Index: 761235 (C4 (Web Text))

Max Activating Token Index: 989

Text #13

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0911. Min Act: -0.0001

Data Index: 1617198 (Python Code)

Max Activating Token Index: 292

Text #14

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0898. Min Act: -0.0001

Data Index: 64755 (C4 (Web Text))

Max Activating Token Index: 13

Text #15

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0906. Min Act: -0.0001

Data Index: 1607227 (Python Code)

Max Activating Token Index: 1019

Text #16

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0907. Min Act: -0.0001

Data Index: 1579014 (Python Code)

Max Activating Token Index: 452

Text #17

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0907. Min Act: -0.0001

Data Index: 519280 (C4 (Web Text))

Max Activating Token Index: 384

Text #18

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0850. Min Act: -0.0001

Data Index: 1064844 (C4 (Web Text))

Max Activating Token Index: 46

Text #19

Max Range: 0.1170. Min Range: -0.1170

Max Act: 0.0864. Min Act: -0.0001

Data Index: 275370 (C4 (Web Text))

Max Activating Token Index: 861