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

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

Neuron 1685 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: 8.6925. Min Range: -8.6925

Max Act: 8.6925. Min Act: -0.1700

Data Index: 1400456 (Python Code)

Max Activating Token Index: 1023

Text #1

Max Range: 8.6925. Min Range: -8.6925

Max Act: 8.5050. Min Act: -0.1700

Data Index: 1370986 (Python Code)

Max Activating Token Index: 1023

Text #2

Max Range: 8.6925. Min Range: -8.6925

Max Act: 8.1925. Min Act: -0.1700

Data Index: 1392214 (Python Code)

Max Activating Token Index: 1023

Text #3

Max Range: 8.6925. Min Range: -8.6925

Max Act: 8.1300. Min Act: -0.1700

Data Index: 1582858 (Python Code)

Max Activating Token Index: 1023

Text #4

Max Range: 8.6925. Min Range: -8.6925

Max Act: 8.0675. Min Act: -0.1700

Data Index: 1652618 (Python Code)

Max Activating Token Index: 1023

Text #5

Max Range: 8.6925. Min Range: -8.6925

Max Act: 8.0987. Min Act: -0.1700

Data Index: 1497591 (Python Code)

Max Activating Token Index: 1023

Text #6

Max Range: 8.6925. Min Range: -8.6925

Max Act: 8.0675. Min Act: -0.1700

Data Index: 1496355 (Python Code)

Max Activating Token Index: 1023

Text #7

Max Range: 8.6925. Min Range: -8.6925

Max Act: 8.0675. Min Act: -0.1700

Data Index: 1504521 (Python Code)

Max Activating Token Index: 1023

Text #8

Max Range: 8.6925. Min Range: -8.6925

Max Act: 8.0050. Min Act: -0.1700

Data Index: 1144636 (C4 (Web Text))

Max Activating Token Index: 1023

Text #9

Max Range: 8.6925. Min Range: -8.6925

Max Act: 8.0675. Min Act: -0.1700

Data Index: 1521568 (Python Code)

Max Activating Token Index: 1023

Text #10

Max Range: 8.6925. Min Range: -8.6925

Max Act: 7.9112. Min Act: -0.1700

Data Index: 686729 (C4 (Web Text))

Max Activating Token Index: 1023

Text #11

Max Range: 8.6925. Min Range: -8.6925

Max Act: 7.8800. Min Act: -0.1700

Data Index: 1508410 (Python Code)

Max Activating Token Index: 1023

Text #12

Max Range: 8.6925. Min Range: -8.6925

Max Act: 7.9112. Min Act: -0.1700

Data Index: 1059094 (C4 (Web Text))

Max Activating Token Index: 1023

Text #13

Max Range: 8.6925. Min Range: -8.6925

Max Act: 7.8800. Min Act: -0.1700

Data Index: 21299 (C4 (Web Text))

Max Activating Token Index: 1023

Text #14

Max Range: 8.6925. Min Range: -8.6925

Max Act: 7.8800. Min Act: -0.1700

Data Index: 1459557 (Python Code)

Max Activating Token Index: 1023

Text #15

Max Range: 8.6925. Min Range: -8.6925

Max Act: 7.9425. Min Act: -0.1700

Data Index: 1495301 (Python Code)

Max Activating Token Index: 1023

Text #16

Max Range: 8.6925. Min Range: -8.6925

Max Act: 7.8800. Min Act: -0.1700

Data Index: 1638319 (Python Code)

Max Activating Token Index: 1023

Text #17

Max Range: 8.6925. Min Range: -8.6925

Max Act: 7.8175. Min Act: -0.1699

Data Index: 1332548 (C4 (Web Text))

Max Activating Token Index: 1023

Text #18

Max Range: 8.6925. Min Range: -8.6925

Max Act: 7.8487. Min Act: -0.1699

Data Index: 666912 (C4 (Web Text))

Max Activating Token Index: 1023

Text #19

Max Range: 8.6925. Min Range: -8.6925

Max Act: 7.8487. Min Act: -0.1700

Data Index: 1227292 (C4 (Web Text))

Max Activating Token Index: 1023