Mayolo Valencia
From Reddit to the Lab
How an online comment opened up a world of possibilities for computer science major Mayolo Valencia
After struggling to land an internship during his sophomore year at Evergreen Valley College in California, Mayolo Valencia took to the Reddit message boards to see if he could find support. There, buried in the deep depths of a thread, Valencia noticed someone mentioning the Community College Internships (CCI) program offered by the Department of Energy Office of Workforce Development for Teachers and Scientists.
“I was looking everywhere possible, and I stumbled upon a random Reddit comment on a post,” Valencia said. “I was like, what is CCI? I applied and after having horrible luck trying to get anything, then boom, suddenly all of these opportunities opened up because of CCI.”
Valencia decided to attend Los Alamos National Laboratory (LANL) to further his research in Artificial Intelligence and machine learning in January 2024 and what was supposed to be a two-month CCI experience, evolved to him staying at LANL for over a year, where he is still a current intern.
Utilizing the resources and classes that come with the CCI program was something Valencia likens to receiving an extra education.
“I grew up in a very low-income setting, so access to computers wasn’t a thing until very, very late in my adolescence,” he shared. “We didn’t have Wi-Fi until I was 15, so I was pretty late to the game, and even then, I didn’t know what I wanted to do until I was 20 or 21.”
Fortunately, while working security at the Amazon Web Services office in Cupertino, Valencia struck up a conversation with an employee who helped him outline how to pursue a career in computer science. This inspired Valencia to return to community college. After taking a machine learning specialization on Coursera, he fell in love with the subject and knew this was his path forward, eventually leading him down the Reddit rabbit hole to CCI.
During his internship, Valencia participated in the 2024 LANL Student Symposium, where his poster presentation on Advanced Feature Learning with Masked Autoencoders for Nuclear Forensics, won first place. Crediting his mentors for helping hone his presentation, Valencia shared he appreciates the network of professionals he works with on a daily basis through the CCI program.
“Being able to reach out to researchers with any question that you have, sets you up to win,” he stated. “I can ask any question and these people who work in the laboratory are some of the most intelligent and well-versed people in the world. It’s wonderful and I love it so much. My mentors have helped me out throughout this whole journey.”
Since pursuing higher education, Valencia’s main focus is understanding how Masked Autoencoders (MAEs) work. He engages with self-supervised learning models that are trained to reconstruct missing information, calling them a “black box,” a reference to how hard these models can be to interpret, specifically how the model makes a prediction. Valencia’s role is to peek inside these black boxes and make the reasoning more transparent.
“I’ve enjoyed this experience in the research setting,” he shared. “It’s been an amazing experience that I would do a million times over. I have been recommending CCI and SULI and all of these other WDTS programs to my friends because it is an experience like no other.”
From a conversation with a kind stranger, to following a lead from Reddit, to now studying at a national laboratory, Valencia learned firsthand how quickly life can move – similar to how the AI landscape is also constantly evolving. Valencia received his associate of science in computer science in May of 2024 and is currently pursuing his bachelor’s at the University of California in Santa Cruz. Afterwards, he hopes to move on to a master’s program and feels the possibilities are endless.
“The freedom to go to conferences and have the name of the lab behind me really helps my ethos when talking to people about the current work I have done,” he explained. “This has been an amazing, amazing opportunity.”
