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

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

Neuron 1534 in Layer 1

Load this data into an Interactive Neuroscope

See Documentation here

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

Text #0

Max Range: 4.1701. Min Range: -4.1701

Max Act: 4.1701. Min Act: -0.1700

Data Index: 1412453 (Python Code)

Max Activating Token Index: 1023

Text #1

Max Range: 4.1701. Min Range: -4.1701

Max Act: 4.0763. Min Act: -0.1700

Data Index: 91943 (C4 (Web Text))

Max Activating Token Index: 1023

Text #2

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.8262. Min Act: -0.1700

Data Index: 1650045 (Python Code)

Max Activating Token Index: 1023

Text #3

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.7949. Min Act: -0.1700

Data Index: 1126049 (C4 (Web Text))

Max Activating Token Index: 1023

Text #4

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.7636. Min Act: -0.1700

Data Index: 79276 (C4 (Web Text))

Max Activating Token Index: 1023

Text #5

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.7636. Min Act: -0.1700

Data Index: 1353490 (C4 (Web Text))

Max Activating Token Index: 1023

Text #6

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.6697. Min Act: -0.1700

Data Index: 457947 (C4 (Web Text))

Max Activating Token Index: 1023

Text #7

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.6697. Min Act: -0.1700

Data Index: 1459932 (Python Code)

Max Activating Token Index: 1023

Text #8

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.6384. Min Act: -0.1700

Data Index: 668385 (C4 (Web Text))

Max Activating Token Index: 1023

Text #9

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.6071. Min Act: -0.1700

Data Index: 1530532 (Python Code)

Max Activating Token Index: 1023

Text #10

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.5758. Min Act: -0.1700

Data Index: 1655049 (Python Code)

Max Activating Token Index: 1023

Text #11

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.6071. Min Act: -0.1700

Data Index: 1473411 (Python Code)

Max Activating Token Index: 1023

Text #12

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.6071. Min Act: -0.1700

Data Index: 917957 (C4 (Web Text))

Max Activating Token Index: 1023

Text #13

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.6071. Min Act: -0.1700

Data Index: 500148 (C4 (Web Text))

Max Activating Token Index: 877

Text #14

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.5445. Min Act: -0.1700

Data Index: 224147 (C4 (Web Text))

Max Activating Token Index: 1023

Text #15

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.5445. Min Act: -0.1700

Data Index: 1454296 (Python Code)

Max Activating Token Index: 1023

Text #16

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.5758. Min Act: -0.1700

Data Index: 785536 (C4 (Web Text))

Max Activating Token Index: 1023

Text #17

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.5445. Min Act: -0.1700

Data Index: 691833 (C4 (Web Text))

Max Activating Token Index: 560

Text #18

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.5758. Min Act: -0.1700

Data Index: 902424 (C4 (Web Text))

Max Activating Token Index: 399

Text #19

Max Range: 4.1701. Min Range: -4.1701

Max Act: 3.6071. Min Act: -0.1700

Data Index: 1147812 (C4 (Web Text))

Max Activating Token Index: 254