Model: GELU Model: 3 Layers, 2048 Neurons per Layer

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

Neuron 1493 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

Transformer Lens Loading: HookedTransformer.from_pretrained('gelu-3l')

Text #0

Max Range: 5.4117. Min Range: -5.4117

Max Act: 5.4117. Min Act: -0.1700

Data Index: 734549 (C4 (Web Text))

Max Activating Token Index: 91

Text #1

Max Range: 5.4117. Min Range: -5.4117

Max Act: 5.3336. Min Act: -0.1700

Data Index: 187751 (C4 (Web Text))

Max Activating Token Index: 644

Text #2

Max Range: 5.4117. Min Range: -5.4117

Max Act: 5.3180. Min Act: -0.1700

Data Index: 896164 (C4 (Web Text))

Max Activating Token Index: 845

Text #3

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.9898. Min Act: -0.1700

Data Index: 1113167 (C4 (Web Text))

Max Activating Token Index: 638

Text #4

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.9742. Min Act: -0.1700

Data Index: 662231 (C4 (Web Text))

Max Activating Token Index: 836

Text #5

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.9430. Min Act: -0.1700

Data Index: 748611 (C4 (Web Text))

Max Activating Token Index: 47

Text #6

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.9117. Min Act: -0.1700

Data Index: 86976 (C4 (Web Text))

Max Activating Token Index: 628

Text #7

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.8961. Min Act: -0.1700

Data Index: 911850 (C4 (Web Text))

Max Activating Token Index: 433

Text #8

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.8492. Min Act: -0.1700

Data Index: 1273077 (C4 (Web Text))

Max Activating Token Index: 54

Text #9

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.8961. Min Act: -0.1700

Data Index: 1330920 (C4 (Web Text))

Max Activating Token Index: 497

Text #10

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.8492. Min Act: -0.1700

Data Index: 1256911 (C4 (Web Text))

Max Activating Token Index: 998

Text #11

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.8180. Min Act: -0.1700

Data Index: 523401 (C4 (Web Text))

Max Activating Token Index: 219

Text #12

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.8180. Min Act: -0.1700

Data Index: 514434 (C4 (Web Text))

Max Activating Token Index: 540

Text #13

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.7867. Min Act: -0.1700

Data Index: 988758 (C4 (Web Text))

Max Activating Token Index: 407

Text #14

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.7711. Min Act: -0.1700

Data Index: 908726 (C4 (Web Text))

Max Activating Token Index: 779

Text #15

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.7867. Min Act: -0.1700

Data Index: 653813 (C4 (Web Text))

Max Activating Token Index: 608

Text #16

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.7867. Min Act: -0.1700

Data Index: 243623 (C4 (Web Text))

Max Activating Token Index: 334

Text #17

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.7711. Min Act: -0.1700

Data Index: 380282 (C4 (Web Text))

Max Activating Token Index: 206

Text #18

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.7242. Min Act: -0.1700

Data Index: 536124 (C4 (Web Text))

Max Activating Token Index: 484

Text #19

Max Range: 5.4117. Min Range: -5.4117

Max Act: 4.6773. Min Act: -0.1700

Data Index: 1555183 (Python Code)

Max Activating Token Index: 930