arXiv subjects

Alice Chen

Publications and source records attributed to Alice Chen.

5 recordsLinked to original sources

arXiv2026

Inferring the Dark from the Observable: Estimating Halo Masses Using Galaxy Properties

The formation and evolution of galaxies are interconnected with that of their host halos. Given this closely related evolutionary history, it should follow that galaxy properties are correlated with their host halo mass. Previous works have shown that star formation rates of galaxies in cluster halos differ systematically from those in field halos, suggesting that host halo mass plays a significant role in the history of galaxy evolution. Here, we examine the correlations between observable galaxy properties - such as stellar mass, star formation rate, half-stellar radius, $r-$band magnitude, and color - and the underlying host halo mass of galaxies. Central galaxy observables are used to probe host halo mass, while satellite galaxy properties are used to estimate subhalo mass. We perform this analysis using random forest, ordinary least squares (OLS), and symbolic regression. Random forest regression can assess the relative significance of different observable galaxy properties, while OLS and symbolic regression can quantify their relationship with the host halo mass. Our results show that gas mass is generally the most significant property in central galaxies of all halo mass ranges, followed by stellar mass and observed $r-$band magnitude. Adding parameters such as the colour and magnitude of galaxies improves host halo mass estimations compared to standard stellar-to-halo mass relations. These results show that halo mass likely has a multi-dimensional dependence on galaxy properties and open the avenue of finding a simple way to constrain halo mass based on what is directly observable in galaxy surveys. Finding the observables that have the strongest correlation with halo mass can also motivate future surveys on which areas to concentrate in detail.

astro-ph.GA↗

arXiv2026

The Minimum Number of Plane Graphs for Sets with Small Hulls

Let $\mathcal{P}$ be a set of $n$ points in $\mathbb{R}^2$, with a convex hull of size $O(n/\log n)$. We prove that $\Omega(12.24^n)$ plane graphs can be drawn on $\mathcal{P}$, the first non-trivial bound for this problem. We also show that a random plane graph, uniformly chosen from the set of all plane graphs of $\mathcal{P}$, has at most $n/12.24$ isolated vertices. This improves upon a previous bound of $n/10.18$. Our analysis is based on studying the expected vertex potentials in a random plane graph. The potential of a vertex is its degree plus the number of vertices visible from it. We show that this quantity can be used to study numbers of plane graphs.

math.CO↗

arXiv2025

Where are all the dark galaxies? Predicting galaxy/halo locations from their bright neighbors

Astronomical objects in our universe that are too faint to be directly detectable exist and are important - an obvious example being dark matter. The same can also apply to very faint baryonic objects, such as low luminosity dwarf galaxies and gravitationally compact objects (e.g., rogue planets, white dwarfs, neutron stars, black holes, dark sirens). While they are very difficult to observe directly, they have locations that are highly important when studying astrophysical phenomena. Here, we use a machine learning algorithm known as symbolic regression to model the probability of a dark object's existence as a function of their separation distances to their closest two ``bright" (directly observable) neighbors, and the distances of these bright objects to each other. An advantage of this algorithm is that it is interpretable by humans and can be used to make reproducible predictions. Galaxies with masses above $10^9 M_{\odot}$ and halos above $10^{12} M_{\odot} $ are the objects that we separate into ``bright" and ``dark" to be used in our analysis. We find that it is possible to predict the density of dark objects using an analytic expression that depends on their distances to their closest bright neighbors in Illustris-TNG galaxy formation simulations, which is significantly better than the (linear) scale-dependent biasing prediction for $k \sim 1.0~ h$Mpc$^{-1}$ (and potentially beyond, if allowed by the resolution). This could potentially open the avenue for finding dark objects based on their vicinity to directly observable bright sources and make future surveys more targeted and efficient.

astro-ph.CO↗

arXiv2023

Interactive Robot Learning from Verbal Correction

The ability to learn and refine behavior after deployment has become ever more important for robots as we design them to operate in unstructured environments like households. In this work, we design a new learning system based on large language model (LLM), OLAF, that allows everyday users to teach a robot using verbal corrections when the robot makes mistakes, e.g., by saying "Stop what you're doing. You should move closer to the cup." A key feature of OLAF is its ability to update the robot's visuomotor neural policy based on the verbal feedback to avoid repeating mistakes in the future. This is in contrast to existing LLM-based robotic systems, which only follow verbal commands or corrections but not learn from them. We demonstrate the efficacy of our design in experiments where a user teaches a robot to perform long-horizon manipulation tasks both in simulation and on physical hardware, achieving on average 20.0% improvement in policy success rate. Videos and more results are at https://ut-austin-rpl.github.io/olaf/

cs.RO↗

arXiv2022

On the Outskirts of Dark Matter Haloes

Halo Models of large scale structure provide powerful and indispensable tools for phenomenological understanding of the clustering of matter in the Universe. While the halo model builds structures out of the superposition of haloes, defining halo profiles in their outskirts - beyond their virial radii - becomes increasingly ambiguous, as one cannot assign matter to individual haloes in a clear way. In this paper, we tackle this question by using numerical N-body simulations to find a systematic definition of mean halo profile that can be extended to large radii. These profiles must be compensated and are the key ingredients for the computation of cosmological correlation functions in the Amended Halo Model. The latter, introduced in our earlier work (arXiv:1912.04872), provides a more physically accurate phenomenological description of nonlinear structure formation, which respects conservation laws on large scales. Here, we extend this model from the matter auto-power spectrum to the halo-matter cross-power spectra, using N-body simulations. We find that the (dimensionless) compensated halo profile, $r^3 \times \rho(r)/M_{200c}$ has a near-universal maximum in the small range $0.03-0.04$ around the virial radius, $r \simeq r_{\rm 200c}$, independent of the halo mass. The profiles cross zero into negative values in the halo outskirts, beyond 2-3$\times r_{\rm 200c}$, consistent with our previous results. We provide preliminary fitting functions for compensated Navarro-Frenk-White (NFW) profiles, and this can be used to compute more physical observables in the Amended Halo Model of large scale structure.

astro-ph.CO↗