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

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

Neuron 1621 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: 3.3989. Min Range: -3.3989

Max Act: 3.3989. Min Act: -0.1700

Data Index: 559649 (C4 (Web Text))

Max Activating Token Index: 942

Text #1

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.3832. Min Act: -0.1700

Data Index: 1278338 (C4 (Web Text))

Max Activating Token Index: 325

Text #2

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2890. Min Act: -0.1700

Data Index: 1447407 (Python Code)

Max Activating Token Index: 942

Text #3

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2733. Min Act: -0.1700

Data Index: 504252 (C4 (Web Text))

Max Activating Token Index: 826

Text #4

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2890. Min Act: -0.1700

Data Index: 1072758 (C4 (Web Text))

Max Activating Token Index: 253

Text #5

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2576. Min Act: -0.1700

Data Index: 219462 (C4 (Web Text))

Max Activating Token Index: 600

Text #6

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2733. Min Act: -0.1700

Data Index: 410368 (C4 (Web Text))

Max Activating Token Index: 808

Text #7

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2418. Min Act: -0.1700

Data Index: 982703 (C4 (Web Text))

Max Activating Token Index: 310

Text #8

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2261. Min Act: -0.1700

Data Index: 1606803 (Python Code)

Max Activating Token Index: 450

Text #9

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2104. Min Act: -0.1700

Data Index: 616659 (C4 (Web Text))

Max Activating Token Index: 385

Text #10

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2418. Min Act: -0.1700

Data Index: 392447 (C4 (Web Text))

Max Activating Token Index: 957

Text #11

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2261. Min Act: -0.1700

Data Index: 121946 (C4 (Web Text))

Max Activating Token Index: 696

Text #12

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2104. Min Act: -0.1700

Data Index: 1337767 (C4 (Web Text))

Max Activating Token Index: 185

Text #13

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2261. Min Act: -0.1700

Data Index: 1157713 (C4 (Web Text))

Max Activating Token Index: 108

Text #14

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.1947. Min Act: -0.1700

Data Index: 156042 (C4 (Web Text))

Max Activating Token Index: 879

Text #15

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2261. Min Act: -0.1700

Data Index: 762860 (C4 (Web Text))

Max Activating Token Index: 30

Text #16

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.1947. Min Act: -0.1700

Data Index: 1246576 (C4 (Web Text))

Max Activating Token Index: 854

Text #17

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.1947. Min Act: -0.1700

Data Index: 1567474 (Python Code)

Max Activating Token Index: 591

Text #18

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.2104. Min Act: -0.1700

Data Index: 1568926 (Python Code)

Max Activating Token Index: 1010

Text #19

Max Range: 3.3989. Min Range: -3.3989

Max Act: 3.1947. Min Act: -0.1700

Data Index: 1563684 (Python Code)

Max Activating Token Index: 1016