By Mackenzie Devereux
Sitton Anne, a physicist at Carnegie Mellon University, has made a groundbreaking discovery that can change the way we understand the universe. Using innovative machine learning techniques, he discovered evidence that four top quarks (4TQ), one of the rarest and heaviest subatomic particles, can be produced simultaneously, at a much higher rate than current physics models predict. His research not only points to the discovery of new particles and forces, but also shows a fundamental shift in the way reality is explored and interpreted.
Of the six members of the Quark Family, Top Quark is the heaviest and rarest. Imagine narrowing the entire gold atom to a point about 300 trillion times smaller than its original size. With more mass, they require more energy to generate more energy and are most sensitive to decaying into other small particles. Furthermore, the mass of these quarks makes them most sensitive to “new physics,” the physical processes that are occurring, but they still don't know or understand. The size of the upper quark is most likely to interact with these mysterious particles and forces. That's why we are very concerned about the production of the four top quarks. This process opens a window for probes in unopened regions of the universe, such as other particles cannot.
These particles and the forces governing them are best explained by the standard model (SM), the basis of modern particle physics. The pinnacle of centuries-old work by countless physicists, SM is the best framework for understanding how particles smaller than atoms act as components of the universe today. SM describes 17 subatomic particles, including the upper quark and its siblings, anti-top quarks. SM has been very successful in explaining particles and predicting interactions, but also has limitations, especially with regard to rare events such as 4TQ production.
Production of 4TQS is much rarer than production of a single topquark pair, which occurs only once with a trillion particle collisions. why? This is because not only one of the top quarks produced, but four top quarks have been created (total of two quarks/antiquark pairs).
SM makes statistical predictions at the frequency to see the production of this special quartet based on what we know about proton annihilation and how other particles interact. Therefore, if the experimental results are very different from what SM predicts, many scientists interpret this as a clue that there may be more in physics than what they currently understand, and that SM is no longer an accurate representation of the whole physical picture.
So, if these particles are very important to understand physics and so on, why are we not already understanding them? This is because searching for 4TQ in currently available data is like searching for four needles in a haystack made of needles. In other words, it's extremely difficult.
Most of the data physicists rely on to study particle collisions, including the production of 4TQ, comes from the world's most powerful particle accelerator, the large hadron crider (LHC). Since operating in 2008, LHC has produced petabytes of data by accelerating protons and their partners (uncreatively named) to 99.999991% of their speed, or 99.999991% per 299,792,458 meters. However, with such a vast dataset, 4TQ events remained elusive.
Top quarks collapse almost instantly, so they are not directly observed. Instead, the way we look for them is to track down particles that come from collapse called jets. However, all of these jet production blends in with the background noise of current data collection capabilities. This makes it extremely challenging to find and distinguish already rare event signals.
This may seem like a dead end to a 4TQ investigation. After all, how can we study what is not detectable? That was up to the recently published Sitton of his doctoral dissertation in experimental high energy physics at Carnegie Mellon University on the subject. Solution to a deadlock? Machine learning and particle physics of bridges to get to places we've never been to before. Machine learning and decision-making models have previously been used to successfully identify single-top quark production. However, this strategy was adapted to specifically identify the 4TQ signal from all background noise. This novel and powerful approach allowed us to predict not only the occurrence of background events, but also the shape of their distribution, giving us a more accurate way to search for noise and distinguish it from actual quark production. Through polyphase training and implementation processes, we expanded the application of AI in physics to solve this near-impossible problem.
So, after all this work, could he detect 4TQ? In fact, he could do more. Using his machine learning strategy, I found the 4TQ signal on runs 2 and 3 in LHC. Two major data collection periods from 2015 to 2018 and 2022 onwards.
During these runs, 4QT production occurs at a rate of 2.7 times higher than SM forecasts, indicating that something else is happening in this process that cannot be explained by current models. But how do you know that this is not just a fluke measurement? To measure importance, physicists use this special scale called the Sigma(σ) scale. Getting 5 on this scale is like getting 1600 in the SAT. You are allowed to hit the jackpot and officially declare your discovery. The reason why physicists can make such a noble claim at 5σ is to represent only one in 3.5 million people who have the results that result from random fluctuations or statistical noise. 4σ means that you can argue that there is strong evidence, but not sufficient to argue that your findings are. An's Sigma value was 3.95σ, which was very close to the finding threshold. In particular, when compared to the predicted value of 1.65σ, the results strongly refer to something large.
What is happening besides the SM that can cause these important deviations in experimental results from predicted values? In short, physicists have not reached broad consensus on what goes beyond current theories. Some physicists have assumed that there are undiscovered particles lurking under our detection. This contradicts particle physics, string theory, and how we understand the universe, but more and more physicists suggest that this mysterious particle may actually exist.
The discovery of 4TQ production at rates far beyond prediction is not merely a statistical anomaly. It is a deep clue pointing to a new chapter in particle physics. Not only does the discovery of AN allow for better detection and study of the four topquark production, but it also illustrates how artificial intelligence shapes the way science is implemented. The growing prevalence of medical imaging, renewable energy, drug discovery, and current particle physics is just the beginning of applying machine learning to further scientific discovery. So, what's next for quarks and particle physics? No one knows. Are there any mysterious phantom particles that we haven't discovered yet, interactions of unknown forces, or anything beyond what we can imagine? Thanks to something like a physicist, we may just know.
