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

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

Neuron 1235 in Layer 2

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.9646. Min Act: -0.1699

Data Index: 1471949 (Python Code)

Max Activating Token Index: 737

Text #1

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.7457. Min Act: -0.1700

Data Index: 684661 (C4 (Web Text))

Max Activating Token Index: 1023

Text #2

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.7457. Min Act: -0.1700

Data Index: 764107 (C4 (Web Text))

Max Activating Token Index: 1023

Text #3

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.6831. Min Act: -0.1700

Data Index: 860209 (C4 (Web Text))

Max Activating Token Index: 1023

Text #4

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.6518. Min Act: -0.1700

Data Index: 1098118 (C4 (Web Text))

Max Activating Token Index: 1023

Text #5

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.5892. Min Act: -0.1699

Data Index: 1655102 (Python Code)

Max Activating Token Index: 1023

Text #6

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.6518. Min Act: -0.1700

Data Index: 1498092 (Python Code)

Max Activating Token Index: 679

Text #7

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.6205. Min Act: -0.1700

Data Index: 1600522 (Python Code)

Max Activating Token Index: 856

Text #8

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.6518. Min Act: -0.1700

Data Index: 757150 (C4 (Web Text))

Max Activating Token Index: 1023

Text #9

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.5578. Min Act: -0.1699

Data Index: 1605588 (Python Code)

Max Activating Token Index: 463

Text #10

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.5892. Min Act: -0.1700

Data Index: 1186713 (C4 (Web Text))

Max Activating Token Index: 1023

Text #11

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.5892. Min Act: -0.1700

Data Index: 1199058 (C4 (Web Text))

Max Activating Token Index: 1023

Text #12

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.5578. Min Act: -0.1700

Data Index: 1012598 (C4 (Web Text))

Max Activating Token Index: 1023

Text #13

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.5265. Min Act: -0.1700

Data Index: 1385036 (Python Code)

Max Activating Token Index: 950

Text #14

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.4952. Min Act: -0.1699

Data Index: 1420277 (Python Code)

Max Activating Token Index: 729

Text #15

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.5265. Min Act: -0.1700

Data Index: 897173 (C4 (Web Text))

Max Activating Token Index: 1023

Text #16

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.5265. Min Act: -0.1700

Data Index: 658334 (C4 (Web Text))

Max Activating Token Index: 1023

Text #17

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.5578. Min Act: -0.1700

Data Index: 128443 (C4 (Web Text))

Max Activating Token Index: 1023

Text #18

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.4952. Min Act: -0.1700

Data Index: 875966 (C4 (Web Text))

Max Activating Token Index: 1023

Text #19

Max Range: 3.9646. Min Range: -3.9646

Max Act: 3.4952. Min Act: -0.1700

Data Index: 777873 (C4 (Web Text))

Max Activating Token Index: 1023