Ph.D. candidate in Computer Science · UCLA
Fred Xu
I'm a Ph.D. candidate in Computer Science at UCLA, researching probabilistic generative modeling, graph learning and uncertainty quantification, with applications to finance and science. Here I write about philosophy, about AI, about film, and about my research.
Philosophy
All philosophy essays →
Philosophy
The Geometry of LoveAnne Carson draws desire as a triangle whose third point holds lover and beloved apart. Following her back to Sappho and Plato, this essay traces the shapes love draws in space and in time, and asks whether there is a love that is neither lack nor pure gift.
Philosophy
Illusions and IllusionsThe Diamond Sutra cuts through illusion twice: once through the forms a mind takes to be solid, beginning with itself, and once more through the illusion of having seen through them.
Philosophy
The Ontology of LoveWhat is this thing, love, that it binds two creatures to each other in life and in death? More fundamentally, what is being-in-love, and how is it related to being-in-the-world?
Philosophy
The Philosophy of BoxingWhy would anyone choose to be hit in the head? Freud's death drive, Musashi's sword after the wars, and what boxing taught me about discipline, fear and flow.
Philosophy
Reflection on my First Boxing Bout - A Warrior's JourneyWhat I learned from losing my first boxing match: sacrifice, the question of why, a sunrise on a trail, and fighting as a spiritual practice.
Philosophy
Jim Simons, A TributeA tribute to Jim Simons, mathematician, codebreaker, investor and patron of science, and an inspiration of mine: on beauty, usefulness and what wealth can do for curiosity.
Philosophy
Coffee Shop ReflectionsWhy I work best among strangers in a coffee shop, and whether the Paris café of Hemingway and the existentialists is a tradition or a fantasy sold by the cup.AI
All AI essays →
AI
A Unified Uncertainty Representation for Graph Neural NetworksDSS-GNN stores each node's embedding as a short polynomial chaos expansion in one shared Gaussian variable and filters it along two spectral axes, graph frequency and chaos order. Readouts of this one representation, computed in a single deterministic pass, give calibrated predictions, OOD scores and robustness to distribution shift.
AI
Graph Anomaly Detection as Finite-Horizon ControlEB-GAD scores graph anomalies with no labels and no training: empirical Bayes fits a graph-aware Gaussian model of normal in closed form, and each node is scored by the finite-horizon control energy needed to steer a neutral start to it under a graph Ornstein–Uhlenbeck process. Best or tied-best AUROC on 9 of 11 benchmarks.
AI
Improving Function Space Flow Matching with Kernel Optimal TransportkFFM replaces the random noise–data pairing of Functional Flow Matching with entropic optimal transport under a kernel-induced cost, the coupling behind the Hilbert Sinkhorn Divergence: a signature kernel for paths, a Sobolev or Euclidean RBF for fields. Same network, same loss, 5–7% more training time, and better samples on eight time-series and PDE benchmarks.
AI
The Virtue of FrictionSome friction only wastes our time, and some is how understanding is built and how hard work becomes absorbing. On learning, skill and flow when a machine can do the work.Film
All films →
Film
The Strong, Silent TypeOdysseus and Machiavelli's prince use silence as a lever, a strategy kept for enemies and rivals. The screen's strong, silent man makes it his default way of feeling instead, a performance that always needs someone to watch it, which is why the camera suits him and why self-help can sell his calm back to men.
Film
The Hermeneutics of Samurai Culture in Film and TelevisionBefore my first boxing match I read Hagakure. Reading it again through Kurosawa, Kobayashi, Mishima, Jarmusch and Samurai Champloo, I find in the samurai ideal a modern invention and a masculine death drive, and in the best films its undoing.
Film
Kwaidan: Strange Tales of Nature and WarMasaki Kobayashi's Kwaidan is usually filed under horror. Seen on a big screen, its four ghost stories read better as mysteries: about a nature full of spirits, and about the dead of a lost war.Research & Papers
All research →Selected papers
NeurIPS 2026
NeurIPS 2026
NeurIPS 2026
KDD 2026
NeurIPS 2025
Findings of ACL 2024