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

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

Neuron 318 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: 3.3725. Min Range: -3.3725

Max Act: 3.3725. Min Act: -0.1700

Data Index: 929002 (C4 (Web Text))

Max Activating Token Index: 995

Text #1

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.3411. Min Act: -0.1700

Data Index: 890808 (C4 (Web Text))

Max Activating Token Index: 920

Text #2

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.1997. Min Act: -0.1700

Data Index: 1612302 (Python Code)

Max Activating Token Index: 969

Text #3

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.1997. Min Act: -0.1700

Data Index: 1068690 (C4 (Web Text))

Max Activating Token Index: 990

Text #4

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.1997. Min Act: -0.1700

Data Index: 353040 (C4 (Web Text))

Max Activating Token Index: 989

Text #5

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.1210. Min Act: -0.1700

Data Index: 181636 (C4 (Web Text))

Max Activating Token Index: 1015

Text #6

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.1525. Min Act: -0.1700

Data Index: 1619536 (Python Code)

Max Activating Token Index: 1002

Text #7

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.0894. Min Act: -0.1700

Data Index: 952176 (C4 (Web Text))

Max Activating Token Index: 1019

Text #8

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.1052. Min Act: -0.1700

Data Index: 1596032 (Python Code)

Max Activating Token Index: 997

Text #9

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.1210. Min Act: -0.1700

Data Index: 1497466 (Python Code)

Max Activating Token Index: 1022

Text #10

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.0894. Min Act: -0.1700

Data Index: 1415141 (Python Code)

Max Activating Token Index: 993

Text #11

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.1210. Min Act: -0.1700

Data Index: 1279240 (C4 (Web Text))

Max Activating Token Index: 1002

Text #12

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.1052. Min Act: -0.1700

Data Index: 449804 (C4 (Web Text))

Max Activating Token Index: 1023

Text #13

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.0894. Min Act: -0.1700

Data Index: 427317 (C4 (Web Text))

Max Activating Token Index: 982

Text #14

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.0737. Min Act: -0.1700

Data Index: 342229 (C4 (Web Text))

Max Activating Token Index: 1023

Text #15

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.0737. Min Act: -0.1700

Data Index: 201615 (C4 (Web Text))

Max Activating Token Index: 983

Text #16

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.0737. Min Act: -0.1700

Data Index: 616476 (C4 (Web Text))

Max Activating Token Index: 941

Text #17

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.0737. Min Act: -0.1700

Data Index: 622923 (C4 (Web Text))

Max Activating Token Index: 927

Text #18

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.0737. Min Act: -0.1700

Data Index: 1625439 (Python Code)

Max Activating Token Index: 927

Text #19

Max Range: 3.3725. Min Range: -3.3725

Max Act: 3.0737. Min Act: -0.1700

Data Index: 697723 (C4 (Web Text))

Max Activating Token Index: 978