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

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

Neuron 1625 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: 3.2309. Min Range: -3.2309

Max Act: 3.2309. Min Act: -0.1700

Data Index: 1001604 (C4 (Web Text))

Max Activating Token Index: 857

Text #1

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.2152. Min Act: -0.1700

Data Index: 1186979 (C4 (Web Text))

Max Activating Token Index: 333

Text #2

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1995. Min Act: -0.1700

Data Index: 1477504 (Python Code)

Max Activating Token Index: 10

Text #3

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1837. Min Act: -0.1700

Data Index: 155852 (C4 (Web Text))

Max Activating Token Index: 723

Text #4

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1837. Min Act: -0.1700

Data Index: 1563619 (Python Code)

Max Activating Token Index: 402

Text #5

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1837. Min Act: -0.1700

Data Index: 1615668 (Python Code)

Max Activating Token Index: 1004

Text #6

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1522. Min Act: -0.1700

Data Index: 935345 (C4 (Web Text))

Max Activating Token Index: 468

Text #7

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1680. Min Act: -0.1700

Data Index: 1457555 (Python Code)

Max Activating Token Index: 470

Text #8

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1522. Min Act: -0.1700

Data Index: 1456818 (Python Code)

Max Activating Token Index: 306

Text #9

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1522. Min Act: -0.1700

Data Index: 1421676 (Python Code)

Max Activating Token Index: 432

Text #10

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1365. Min Act: -0.1700

Data Index: 1581716 (Python Code)

Max Activating Token Index: 6

Text #11

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1522. Min Act: -0.1700

Data Index: 1402879 (Python Code)

Max Activating Token Index: 541

Text #12

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1365. Min Act: -0.1700

Data Index: 1508564 (Python Code)

Max Activating Token Index: 431

Text #13

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1522. Min Act: -0.1700

Data Index: 1538180 (Python Code)

Max Activating Token Index: 8

Text #14

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1365. Min Act: -0.1700

Data Index: 1585098 (Python Code)

Max Activating Token Index: 6

Text #15

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1365. Min Act: -0.1700

Data Index: 1600845 (Python Code)

Max Activating Token Index: 185

Text #16

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1365. Min Act: -0.1700

Data Index: 1418679 (Python Code)

Max Activating Token Index: 215

Text #17

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1365. Min Act: -0.1700

Data Index: 1643005 (Python Code)

Max Activating Token Index: 285

Text #18

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1522. Min Act: -0.1700

Data Index: 1479750 (Python Code)

Max Activating Token Index: 610

Text #19

Max Range: 3.2309. Min Range: -3.2309

Max Act: 3.1522. Min Act: -0.1700

Data Index: 963039 (C4 (Web Text))

Max Activating Token Index: 242