"of" in fixed name phrases

What this neuron does

This neuron lights up on the word "of" in set phrases and names like "City of David", "Son of God" or "Land of 10,000 Lakes". When it was turned up, the model's text drifted into unrelated topics and did not clearly change in a single direction.

Written by Claude from this neuron's activations and steering outputs. Confidence: medium. Think it is wrong? Flag it (opens a prefilled issue on GitHub).

Fires most on

The text windows where this neuron's activation was highest. Amber marks positive activation; the outlined token is the peak. Hover a token for its strength.

1. #1 peak

   great joy that will be for all the people. 11 For unto you is born this day in the city of David a Savior, who is Christ the Lord. 12 And this will

2. #2 peak

   was exchanged for money will come back to him in the year of Jubilee. For the houses of the towns of the Levites are their property among the children of Israel.⏎But

3. #3 peak

   away with the impression that Hindu and Muslim beliefs are utterly wrong, and certainly the basic premise of Islam condemns the heretical Christian claim that Jesus was the son of God. In 1987, it boasted a membership of more than

4. #4 peak

   for the sign of the Prophet Jonah, “As Jonah was in the belly of the whale three days and three nights, so also shall the Son of Man be

5. #5 peak

   lifting as you pull. The experience when it happens can be anything from an element that just won’t turn on to something resembling a Fourth of July fireworks display. This means your oven and stove will be out of commission until

6. #6 peak

   858, created from the eastern half of the Minnesota Territory. The declare has a large number of lakes, and is known by the slogan the “Land of 10,000 Lakes”. Its certified motto is L’to

Steering

Push this neuron up or down while the model writes, and watch what changes. The outputs were generated ahead of time, so nothing runs in your browser.

Most negative on

Where the activation went furthest below zero. Blue marks negative activation. Many neurons mean something different in each direction.

1. #1 peak

   a...⏎Birthplace: Shadwell, Albemarle County, VirginiaReligion: Christian (no specific denomination)College Attended: College of William and Mary,...⏎Spend a day in the life of the multi-talented man

2. #2 peak

   survey courses in British, European, and World History.⏎Alan J. Ward is the Class of 1935 Professor of Government at the College of William and Mary in Virginia. He is a past president of the American Conference for Irish Studies

3. #3 peak

   .⏎Read the rest of this entry »⏎Filed under: Gaming, Services, Software, 3 million dollar grant, Associate Professor, Augmented reality, Bill & Melinda Gates Foundation, Common Core, Common Core standards, Education, education arcade

4. #4 peak

   ⏎$3 million grant from the Bill & Melinda Gates Foundation will fund development⏎by Eric Klopfer⏎MIT Education

Which neurons in earlier layers feed this one, and which later ones it feeds. Amber links excite, blue links suppress; longer bars are stronger.

Activation histogram

How often this neuron fires at each strength. Almost always it does close to nothing (the tall bars in the middle); the snippets above are the rare moments it fired hard, amber for positive and blue for negative.

Logit lens

The words this neuron nudges the model toward (amber) or away from (blue) when it fires. A hint, not a label: many neurons work through later layers instead.