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

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

Neuron 1030 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: 3.1637. Min Range: -3.1637

Max Act: 3.1637. Min Act: -0.1700

Data Index: 1595198 (Python Code)

Max Activating Token Index: 500

Text #1

Max Range: 3.1637. Min Range: -3.1637

Max Act: 3.1479. Min Act: -0.1700

Data Index: 367567 (C4 (Web Text))

Max Activating Token Index: 198

Text #2

Max Range: 3.1637. Min Range: -3.1637

Max Act: 3.0691. Min Act: -0.1700

Data Index: 1537387 (Python Code)

Max Activating Token Index: 810

Text #3

Max Range: 3.1637. Min Range: -3.1637

Max Act: 3.0059. Min Act: -0.1700

Data Index: 1329766 (C4 (Web Text))

Max Activating Token Index: 909

Text #4

Max Range: 3.1637. Min Range: -3.1637

Max Act: 3.0533. Min Act: -0.1700

Data Index: 1483170 (Python Code)

Max Activating Token Index: 342

Text #5

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.9743. Min Act: -0.1700

Data Index: 1549659 (Python Code)

Max Activating Token Index: 214

Text #6

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.9743. Min Act: -0.1700

Data Index: 1402567 (Python Code)

Max Activating Token Index: 351

Text #7

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.9901. Min Act: -0.1700

Data Index: 1363241 (Python Code)

Max Activating Token Index: 510

Text #8

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.9268. Min Act: -0.1700

Data Index: 1353507 (C4 (Web Text))

Max Activating Token Index: 937

Text #9

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.9268. Min Act: -0.1700

Data Index: 892792 (C4 (Web Text))

Max Activating Token Index: 461

Text #10

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.9268. Min Act: -0.1700

Data Index: 1592046 (Python Code)

Max Activating Token Index: 439

Text #11

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.9585. Min Act: -0.1700

Data Index: 1363945 (Python Code)

Max Activating Token Index: 399

Text #12

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.9268. Min Act: -0.1700

Data Index: 1392184 (Python Code)

Max Activating Token Index: 958

Text #13

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.9110. Min Act: -0.1700

Data Index: 1528577 (Python Code)

Max Activating Token Index: 578

Text #14

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.8951. Min Act: -0.1700

Data Index: 1437918 (Python Code)

Max Activating Token Index: 485

Text #15

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.9743. Min Act: -0.1700

Data Index: 444421 (C4 (Web Text))

Max Activating Token Index: 500

Text #16

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.8792. Min Act: -0.1700

Data Index: 1606184 (Python Code)

Max Activating Token Index: 318

Text #17

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.9585. Min Act: -0.1700

Data Index: 1552210 (Python Code)

Max Activating Token Index: 298

Text #18

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.8792. Min Act: -0.1700

Data Index: 1375922 (Python Code)

Max Activating Token Index: 470

Text #19

Max Range: 3.1637. Min Range: -3.1637

Max Act: 2.8951. Min Act: -0.1700

Data Index: 1579884 (Python Code)

Max Activating Token Index: 152