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

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

Neuron 496 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 4.2543. Min Range: -4.2543

Max Act: 4.2543. Min Act: -0.1700

Data Index: 1448843 (Python Code)

Max Activating Token Index: 1023

Text #1

Max Range: 4.2543. Min Range: -4.2543

Max Act: 4.1918. Min Act: -0.1700

Data Index: 1623376 (Python Code)

Max Activating Token Index: 1023

Text #2

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.9104. Min Act: -0.1700

Data Index: 1638030 (Python Code)

Max Activating Token Index: 1023

Text #3

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.8479. Min Act: -0.1700

Data Index: 57610 (C4 (Web Text))

Max Activating Token Index: 1023

Text #4

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.8792. Min Act: -0.1700

Data Index: 96237 (C4 (Web Text))

Max Activating Token Index: 1023

Text #5

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.8479. Min Act: -0.1699

Data Index: 757913 (C4 (Web Text))

Max Activating Token Index: 1023

Text #6

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.8166. Min Act: -0.1700

Data Index: 1594157 (Python Code)

Max Activating Token Index: 1023

Text #7

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.7540. Min Act: -0.1700

Data Index: 1145858 (C4 (Web Text))

Max Activating Token Index: 1023

Text #8

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.8166. Min Act: -0.1700

Data Index: 1375686 (Python Code)

Max Activating Token Index: 1023

Text #9

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.7853. Min Act: -0.1700

Data Index: 1361430 (Python Code)

Max Activating Token Index: 1023

Text #10

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.8166. Min Act: -0.1700

Data Index: 1463190 (Python Code)

Max Activating Token Index: 1023

Text #11

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.7540. Min Act: -0.1700

Data Index: 1479817 (Python Code)

Max Activating Token Index: 1023

Text #12

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.7227. Min Act: -0.1700

Data Index: 1520392 (Python Code)

Max Activating Token Index: 1023

Text #13

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.7227. Min Act: -0.1700

Data Index: 1494165 (Python Code)

Max Activating Token Index: 1023

Text #14

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.5349. Min Act: -0.1700

Data Index: 1515108 (Python Code)

Max Activating Token Index: 492

Text #15

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.6915. Min Act: -0.1700

Data Index: 1575808 (Python Code)

Max Activating Token Index: 418

Text #16

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.6602. Min Act: -0.1700

Data Index: 1533309 (Python Code)

Max Activating Token Index: 1023

Text #17

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.6602. Min Act: -0.1700

Data Index: 1413357 (Python Code)

Max Activating Token Index: 1023

Text #18

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.6288. Min Act: -0.1700

Data Index: 1529312 (Python Code)

Max Activating Token Index: 590

Text #19

Max Range: 4.2543. Min Range: -4.2543

Max Act: 3.5662. Min Act: -0.1700

Data Index: 1360936 (Python Code)

Max Activating Token Index: 330