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

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

Neuron 1753 in Layer 0

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 8.4073. Min Range: -8.4073

Max Act: 8.4073. Min Act: -0.1700

Data Index: 1098961 (C4 (Web Text))

Max Activating Token Index: 29

Text #1

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.9073. Min Act: -0.1700

Data Index: 419608 (C4 (Web Text))

Max Activating Token Index: 92

Text #2

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.8760. Min Act: -0.1700

Data Index: 782005 (C4 (Web Text))

Max Activating Token Index: 27

Text #3

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.5948. Min Act: -0.1700

Data Index: 289959 (C4 (Web Text))

Max Activating Token Index: 132

Text #4

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.5948. Min Act: -0.1700

Data Index: 1215410 (C4 (Web Text))

Max Activating Token Index: 28

Text #5

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.4073. Min Act: -0.1700

Data Index: 523396 (C4 (Web Text))

Max Activating Token Index: 114

Text #6

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.3448. Min Act: -0.1700

Data Index: 1377144 (Python Code)

Max Activating Token Index: 20

Text #7

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.3760. Min Act: -0.1700

Data Index: 1098900 (C4 (Web Text))

Max Activating Token Index: 1023

Text #8

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.3448. Min Act: -0.1700

Data Index: 1083730 (C4 (Web Text))

Max Activating Token Index: 20

Text #9

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.3448. Min Act: -0.1700

Data Index: 1350704 (C4 (Web Text))

Max Activating Token Index: 99

Text #10

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.3760. Min Act: -0.1700

Data Index: 1268574 (C4 (Web Text))

Max Activating Token Index: 198

Text #11

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.3448. Min Act: -0.1700

Data Index: 1199466 (C4 (Web Text))

Max Activating Token Index: 39

Text #12

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.2510. Min Act: -0.1700

Data Index: 55157 (C4 (Web Text))

Max Activating Token Index: 8

Text #13

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.2198. Min Act: -0.1700

Data Index: 384008 (C4 (Web Text))

Max Activating Token Index: 1023

Text #14

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.2823. Min Act: -0.1700

Data Index: 80221 (C4 (Web Text))

Max Activating Token Index: 36

Text #15

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.2198. Min Act: -0.1700

Data Index: 735132 (C4 (Web Text))

Max Activating Token Index: 1023

Text #16

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.1885. Min Act: -0.1700

Data Index: 294093 (C4 (Web Text))

Max Activating Token Index: 69

Text #17

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.1885. Min Act: -0.1700

Data Index: 123776 (C4 (Web Text))

Max Activating Token Index: 1023

Text #18

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.1885. Min Act: -0.1700

Data Index: 719166 (C4 (Web Text))

Max Activating Token Index: 20

Text #19

Max Range: 8.4073. Min Range: -8.4073

Max Act: 7.1260. Min Act: -0.1700

Data Index: 1586533 (Python Code)

Max Activating Token Index: 39