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

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

Neuron 788 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: 3.8598. Min Range: -3.8598

Max Act: 3.8598. Min Act: -0.1700

Data Index: 1287981 (C4 (Web Text))

Max Activating Token Index: 296

Text #1

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.6564. Min Act: -0.1700

Data Index: 1609022 (Python Code)

Max Activating Token Index: 208

Text #2

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.6251. Min Act: -0.1700

Data Index: 1610361 (Python Code)

Max Activating Token Index: 358

Text #3

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5938. Min Act: -0.1700

Data Index: 261445 (C4 (Web Text))

Max Activating Token Index: 629

Text #4

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5625. Min Act: -0.1700

Data Index: 647755 (C4 (Web Text))

Max Activating Token Index: 619

Text #5

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5938. Min Act: -0.1700

Data Index: 811058 (C4 (Web Text))

Max Activating Token Index: 499

Text #6

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5782. Min Act: -0.1700

Data Index: 146193 (C4 (Web Text))

Max Activating Token Index: 533

Text #7

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5625. Min Act: -0.1700

Data Index: 234758 (C4 (Web Text))

Max Activating Token Index: 928

Text #8

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5625. Min Act: -0.1700

Data Index: 1632547 (Python Code)

Max Activating Token Index: 35

Text #9

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5469. Min Act: -0.1700

Data Index: 1649129 (Python Code)

Max Activating Token Index: 750

Text #10

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5469. Min Act: -0.1700

Data Index: 1224623 (C4 (Web Text))

Max Activating Token Index: 631

Text #11

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5469. Min Act: -0.1700

Data Index: 697787 (C4 (Web Text))

Max Activating Token Index: 807

Text #12

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5312. Min Act: -0.1700

Data Index: 1595402 (Python Code)

Max Activating Token Index: 798

Text #13

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5469. Min Act: -0.1700

Data Index: 1641693 (Python Code)

Max Activating Token Index: 97

Text #14

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5312. Min Act: -0.1700

Data Index: 829313 (C4 (Web Text))

Max Activating Token Index: 564

Text #15

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5312. Min Act: -0.1700

Data Index: 1621862 (Python Code)

Max Activating Token Index: 193

Text #16

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5155. Min Act: -0.1700

Data Index: 1568947 (Python Code)

Max Activating Token Index: 714

Text #17

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.4999. Min Act: -0.1700

Data Index: 485131 (C4 (Web Text))

Max Activating Token Index: 237

Text #18

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.5155. Min Act: -0.1700

Data Index: 168237 (C4 (Web Text))

Max Activating Token Index: 780

Text #19

Max Range: 3.8598. Min Range: -3.8598

Max Act: 3.4842. Min Act: -0.1700

Data Index: 1525798 (Python Code)

Max Activating Token Index: 183