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

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

Neuron 1005 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: 3.5921. Min Range: -3.5921

Max Act: 3.5921. Min Act: -0.1700

Data Index: 1388010 (Python Code)

Max Activating Token Index: 8

Text #1

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.5295. Min Act: -0.1700

Data Index: 1600976 (Python Code)

Max Activating Token Index: 8

Text #2

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.5138. Min Act: -0.1700

Data Index: 1628792 (Python Code)

Max Activating Token Index: 10

Text #3

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.5138. Min Act: -0.1700

Data Index: 1532578 (Python Code)

Max Activating Token Index: 10

Text #4

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4825. Min Act: -0.1700

Data Index: 1572317 (Python Code)

Max Activating Token Index: 9

Text #5

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4825. Min Act: -0.1700

Data Index: 1389289 (Python Code)

Max Activating Token Index: 9

Text #6

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4982. Min Act: -0.1700

Data Index: 1611321 (Python Code)

Max Activating Token Index: 7

Text #7

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4825. Min Act: -0.1700

Data Index: 1625676 (Python Code)

Max Activating Token Index: 9

Text #8

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4825. Min Act: -0.1700

Data Index: 1366528 (Python Code)

Max Activating Token Index: 9

Text #9

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4825. Min Act: -0.1700

Data Index: 1430135 (Python Code)

Max Activating Token Index: 10

Text #10

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4825. Min Act: -0.1700

Data Index: 1379919 (Python Code)

Max Activating Token Index: 13

Text #11

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4668. Min Act: -0.1700

Data Index: 1560274 (Python Code)

Max Activating Token Index: 11

Text #12

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4668. Min Act: -0.1700

Data Index: 1430044 (Python Code)

Max Activating Token Index: 13

Text #13

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4668. Min Act: -0.1700

Data Index: 1402617 (Python Code)

Max Activating Token Index: 11

Text #14

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4511. Min Act: -0.1700

Data Index: 1384476 (Python Code)

Max Activating Token Index: 12

Text #15

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4511. Min Act: -0.1700

Data Index: 1467231 (Python Code)

Max Activating Token Index: 12

Text #16

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4511. Min Act: -0.1700

Data Index: 1604006 (Python Code)

Max Activating Token Index: 18

Text #17

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4355. Min Act: -0.1700

Data Index: 1530602 (Python Code)

Max Activating Token Index: 6

Text #18

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4355. Min Act: -0.1700

Data Index: 1364040 (Python Code)

Max Activating Token Index: 6

Text #19

Max Range: 3.5921. Min Range: -3.5921

Max Act: 3.4355. Min Act: -0.1700

Data Index: 1614216 (Python Code)

Max Activating Token Index: 21