(2026-11-17) The conformal window of QCD appears when we increase the number of quarks. How the IR changes from the confined phase to a conformal phase is a long-standing question in QCD. In a recent work, we test the so-called "fixed point merger" scenario by computing the anomalous dimension of a four-quark operator in perturbation theory. I gave a talk about this work at ICTP once; here are the slides.
There is an ancient Indian story, “The Blind and the Elephant,” from the Nirvana Sutra (大乘大般涅槃经). Doing research is sometimes very much like that story. We try to study a problem using all available methods and perspectives, hoping to reach consensus. Our perturbative calculation is one perspective. When Tom did this heroic calculation, he did not use AI. This is "the twilight of the Heroic Age".
(2026-11-17) In a recent paper, we proposed an algorithm that starts from a known two-dimensional conformal field theory and generates a web of VOAs with the same central charge (see Section 2). It is not guaranteed that the algorithm generates the complete web. It is, however, very powerful. It will be useful for the CFT zoo project. In the future, we wish to build an Atlas of CFTs with low central charges, starting from known seeds from coset construction, parafermions, tensor product of minimal models, N=1 minimal models, and W_N minimal models. At current AI4S age, this project becomes manageable. The left is a web at c=5/3 with rich structure.
When applied to a tensor-product CFT, the web also allows us to study conformal interfaces between the two theories.
(2026-10-27) Can we make a critical state of matter in the lab? The recent rapid development in the use of a Rydberg atom array to study quantum many-body systems offers the opportunity to directly create a critical quantum state in the lab and test conformal symmetry. In [ArXiv: 2402.15376], experimentalists have created a state whose correlation length is 13 atoms. In our new paper, we proposed that optimizing the ramp profile for adiabatic evolution can yield a longer correlation length in a shorter evolution time. We argue that using Rydberg atoms to create a good quantum critical state is feasible with current technology.
Here is a nice summary from gist.science.
(2026-07-05) When two conformal field theories meet, they create a so-called conformal interface. Among the conformal interfaces, there is a special type called renormalization group (RG) interfaces. In a recent paper [arXiv: 2605.23848], we constructed on the lattice the RG interface between the tricritical Ising CFT and the Ising CFT fully analytically.
Recently, the realization of 1+1D critical phases of matter using Rydberg atom arrays is advancing rapidly. The Ising CFT was realized in [1], and the tricritical Ising CFT was realized in [2]. In addition to being interesting in its own right, our newly constructed interface can be realized in these quantum optical systems. Realizing these critical systems is a crucial step in developing and benchmarking the experimental platform and, therefore, in preparing to study more complex quantum systems (such as the Hubbard model) whose exact solution is beyond the reach of any currently available theoretical methods.
(2026-11-17) Today, we posted a paper towards the classification of conformal interfaces between the Ising CFT and the tricritical Ising CFT.
(2025-07-22) Irrational CFTs are CFTs with central charge c>1 and no extended algebra. Using coupled anyon chains, preserving a certain non-invertible symmetry, we got a CFT with c≈1.15. The symmetry and the spectrum do not seem to coincide with any known rational CFTs. This may be the irrational CFT with the smallest central charge. The paper can be found on Arxiv.
(2025-7-31) Our work relied heavily on an anyon chain formulation of the tricritical Ising model. The CFT is obtained by coupling two anyon chains together, triggering an RG flow which preserves a certain non-invertible symmetry. One exciting direction is that the anyon chain formulation works for a generic fusion category, therefore allows us to study RG flow from coupled generic rational CFTs. With this method, we should be able to gain a better understanding of the landscape of 2D CFTs.
(2025-12-1) It turns out that the c≈1.15 critical point that we identified as a CFT is in fact a weakly first-order phase transition. We have uploaded the 2nd version of our paper. The phase diagram of the 2-coupled chain model has a new c=1.77 critical point, which is hard to interpret as a rational conformal field theory. We have also found the conformal phase with c=2.09, corresponding to three copies of the tricritical Ising model coupled via the sigma' operators. This conformal phase is likely to be an irrational CFT.
(2023-06-01) Ordinary phase transitions happen between a high-temperature disordered phase and a low-temperature ordered phase. However, more exotic inverted phase diagrams are possible with the help of certain conformal field theories. See the talk "Spontaneous breaking of finite group symmetries at all temperatures" and the paper [Arxiv].
(2024-09-18) In a recent paper, we studied this problem using functional RG [Arixv], and showed that indeed some quantum field theory models would show persistent order that is unmeltable, building on the genius idea in [Weinberg, 74] and [Chai et al, 2020].
(2024-12-16) In [paper] by Zohar Komargodski and Fedor K. Popov, the authors rigorously established that persistent order exists. This phenomenon may have important applications in many areas of physics. Imagine that we have a new phase of the standard model that persists to an infinite temperature (also called "symmetry non-restoration"), this will change the history of the universe (see, for example [1] [2] [3][4]). If we can construct a model where the BKT phase can persist to an infinite temperature, we get a high-temperature superfluid. It is time to re-investigate the application of persistent order to early universe cosmology and condensed matter physics!
(2025-01-16) Today, a popular science article covering the story of persistent order appeared in the Quanta Magzine. Our work was briefly mentioned.
(2022-10-22) In Monte Carlo simulation, the lattice's size can be interpreted as the scale of a renormalization group flow. Combining the recent results on Cubic CFT with quantum Monte Carlo simulation, we discovered a new phase of the fully packed quantum loop model on a triangular lattice. See my talk "Conformal field theory approach to quantum loop models" and the paper [Arxiv].
(2024-12-03) In [Arxiv], we studied the finite-temperature phase diagram of the Quantum Loop model. The vison-plaquette phase previously discovered is shown to persist to a finite temperature. One can study the finite temperature phase transition by considering putting the conformal field theory on S1*R2. See a recent talk titled "From the Cubic CFT to the Quantum Loop model" that I gave.
(2024-06-15) Recently, I introduced a new perturbation theory to study conformal field theories, which involves performing the well-known long-range perturbation theory and then imposing the local conformal Ward Identity to recover the conformal data of the local CFTs. This method works perfectly for the O(N) vector models in 3D after properly re-summing the long-range perturbation series. [Arxiv]
(2024-09-18) We need to properly re-sum the asymptotic perturbative series to approach the strong coupling region. The large-order asymptotics of the series are controlled by a soliton of a nonlinear equation with fractional Laplacian, which was first written down in a famous paper by E. Lieb. This suggests that the perturbative series is Borel re-summable.
(2025-01-30) These long-range quantum field theories are also important in statistical physics, for the interaction among particles on the lattice is usually long in nature. Here are some Monte Carlo simulations of the quantum Heinsberg models. In particular, the quantum phase transition of the bilayer long-range Heinsberg model is governed by a non-traditional QFT where space and time have different scaling behavior, see the papers [Arxiv_1] [Arxiv_2].
(2023-01-09) Conformal bootstrap is a method to constrain the scaling dimension of CFT operators, when applied to the super-Ising model in 2+1 dimensions, it lets us determine certain conformal data to high precision. For details, check my talk "Bootstrap minimal superconformal field theory in 2+1D" and the papers [Arxiv_1] [Arxiv_2] [Arxiv_3].
(2023-11-16)The large charge section of O(3) satisfies a formula predicted by effective action. We can measure the universal constants in the formula using numerical bootstrap. The data also allows us to perform conformal perturbation to study another important theory called the Cubic CFT. [Arxiv][slides]
(2025-01-30) The word "bootstrap" here means to lift up oneself by pulling one's bootstraps. It is interesting to see this metaphor appearing in different cultures, in ancient China, there is a story of a general (项羽) who is so strengthful that he could sit on a chair and lift the chair (and himself) up. In conformal field theory research, the word "bootstrap" means a method by which one can solve a system purely relying on self-consistency conditions.
(2023-06-09) Using the mathematical result on the classification of finite subgroups of the O(5) Lie group, we classified the stable irreducible perturbative fixed points of five scalars' field theories. See my talk "Towards classifying perturbative fixed point in 4-eps expansion" (or here) and the paper [Arxiv].
(2025-01-30) The classification of perturbative CFTs in 4-epsilon expansion involves solving a set of polynomial equations. The standard way to attack this type of problem is to calculate the so-called "Gröbner basis". It may be possible that a quantum computer can speed up the calculation!
"Conformal field theory approach to (quantum) phase transitions". This is a two-hour lecture I gave during the "Hong Kong Computational and Theoretical Physics Study Group". The video recordings can be found (lecture I, lecture II). Please also take a look at the review/lecture note on Scalar CFTs from lattice systems.
"Introduction to 2+1D conformal field theories." This is a four-hour lecture that I gave at CPhT, Polytechnique, and also HKU. The lecture notes can be found here.