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

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

Neuron 384 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.3554. Min Act: -0.1700

Data Index: 288462 (C4 (Web Text))

Max Activating Token Index: 830

Text #1

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.3240. Min Act: -0.1700

Data Index: 238753 (C4 (Web Text))

Max Activating Token Index: 465

Text #2

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.2140. Min Act: -0.1700

Data Index: 1400687 (Python Code)

Max Activating Token Index: 477

Text #3

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1983. Min Act: -0.1700

Data Index: 427739 (C4 (Web Text))

Max Activating Token Index: 6

Text #4

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1983. Min Act: -0.1700

Data Index: 1130372 (C4 (Web Text))

Max Activating Token Index: 961

Text #5

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.2140. Min Act: -0.1700

Data Index: 1314362 (C4 (Web Text))

Max Activating Token Index: 959

Text #6

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1825. Min Act: -0.1700

Data Index: 1473268 (Python Code)

Max Activating Token Index: 586

Text #7

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1825. Min Act: -0.1700

Data Index: 595788 (C4 (Web Text))

Max Activating Token Index: 680

Text #8

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1510. Min Act: -0.1700

Data Index: 1451256 (Python Code)

Max Activating Token Index: 125

Text #9

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1510. Min Act: -0.1700

Data Index: 760778 (C4 (Web Text))

Max Activating Token Index: 146

Text #10

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1353. Min Act: -0.1700

Data Index: 117246 (C4 (Web Text))

Max Activating Token Index: 12

Text #11

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1353. Min Act: -0.1700

Data Index: 759385 (C4 (Web Text))

Max Activating Token Index: 929

Text #12

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1195. Min Act: -0.1700

Data Index: 285910 (C4 (Web Text))

Max Activating Token Index: 845

Text #13

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1195. Min Act: -0.1700

Data Index: 1607081 (Python Code)

Max Activating Token Index: 411

Text #14

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1353. Min Act: -0.1700

Data Index: 596337 (C4 (Web Text))

Max Activating Token Index: 351

Text #15

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1038. Min Act: -0.1700

Data Index: 822974 (C4 (Web Text))

Max Activating Token Index: 657

Text #16

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.0564. Min Act: -0.1700

Data Index: 821683 (C4 (Web Text))

Max Activating Token Index: 829

Text #17

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1038. Min Act: -0.1700

Data Index: 1060662 (C4 (Web Text))

Max Activating Token Index: 947

Text #18

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.1038. Min Act: -0.1700

Data Index: 1041481 (C4 (Web Text))

Max Activating Token Index: 995

Text #19

Max Range: 3.3554. Min Range: -3.3554

Max Act: 3.0880. Min Act: -0.1700

Data Index: 1037303 (C4 (Web Text))

Max Activating Token Index: 7