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

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

Neuron 1492 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: 3.6936. Min Range: -3.6936

Max Act: 3.6936. Min Act: -0.1700

Data Index: 1492907 (Python Code)

Max Activating Token Index: 378

Text #1

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.5057. Min Act: -0.1700

Data Index: 1649956 (Python Code)

Max Activating Token Index: 660

Text #2

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.5214. Min Act: -0.1700

Data Index: 1468342 (Python Code)

Max Activating Token Index: 972

Text #3

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.4744. Min Act: -0.1700

Data Index: 1128970 (C4 (Web Text))

Max Activating Token Index: 162

Text #4

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.4116. Min Act: -0.1700

Data Index: 1449672 (Python Code)

Max Activating Token Index: 248

Text #5

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.4116. Min Act: -0.1700

Data Index: 1363311 (Python Code)

Max Activating Token Index: 421

Text #6

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3960. Min Act: -0.1700

Data Index: 1483535 (Python Code)

Max Activating Token Index: 672

Text #7

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3803. Min Act: -0.1700

Data Index: 1643976 (Python Code)

Max Activating Token Index: 277

Text #8

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3803. Min Act: -0.1700

Data Index: 1406601 (Python Code)

Max Activating Token Index: 324

Text #9

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3803. Min Act: -0.1700

Data Index: 1504075 (Python Code)

Max Activating Token Index: 758

Text #10

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3646. Min Act: -0.1700

Data Index: 1378901 (Python Code)

Max Activating Token Index: 299

Text #11

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3803. Min Act: -0.1700

Data Index: 1497042 (Python Code)

Max Activating Token Index: 234

Text #12

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3489. Min Act: -0.1700

Data Index: 1471365 (Python Code)

Max Activating Token Index: 280

Text #13

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3332. Min Act: -0.1700

Data Index: 1500936 (Python Code)

Max Activating Token Index: 887

Text #14

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3489. Min Act: -0.1700

Data Index: 1373986 (Python Code)

Max Activating Token Index: 594

Text #15

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3960. Min Act: -0.1700

Data Index: 1422893 (Python Code)

Max Activating Token Index: 509

Text #16

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3489. Min Act: -0.1700

Data Index: 1420095 (Python Code)

Max Activating Token Index: 255

Text #17

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3175. Min Act: -0.1700

Data Index: 1401955 (Python Code)

Max Activating Token Index: 670

Text #18

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3332. Min Act: -0.1700

Data Index: 1595998 (Python Code)

Max Activating Token Index: 728

Text #19

Max Range: 3.6936. Min Range: -3.6936

Max Act: 3.3803. Min Act: -0.1700

Data Index: 1450603 (Python Code)

Max Activating Token Index: 807