共找到 20 条结果
The Steiner Forest problem, also known as the Generalized Steiner Tree problem, is a fundamental optimization problem on edge-weighted graphs where, given a set of vertex pairs, the goal is to select a minimum-cost subgraph such that each pair is connected. This problem generalizes the Steiner Tree problem, first introduced in 1811, for which the best known approximation factor is 1.39 [Byrka, Grandoni, Rothvoß, and Sanità, 2010] (Best Paper award, STOC 2010). The celebrated work of [Agrawal, Klein, and Ravi, 1989] (30-Year Test-of-Time award, STOC 2023), along with refinements by [Goemans and Williamson, 1992] (SICOMP'95), established a 2-approximation for Steiner Forest over 35 years ago. Jain's (FOCS'98) pioneering iterative rounding techniques later extended these results to higher connectivity settings. Despite the long-standing importance of this problem, breaking the approximation factor of 2 has remained a major challenge, raising suspicions that achieving a better factor -- similar to Vertex Cover -- might indeed be hard. Notably, fundamental works, including those by Gupta and Kumar (STOC'15) and Groß et al. (ITCS'18), introduced 96- and 69-approximation algorithms, possi
We present photometric and spectroscopic observations of SN 2023fyq, a type Ibn supernova in the nearby galaxy NGC 4388 (D$\simeq$18~Mpc). In addition, we trace long-standing precursor emission at the position of SN 2023fyq using data from DLT40, ATLAS, ZTF, ASAS-SN, Swift, and amateur astronomer Koichi Itagaki. Precursor activity is observed up to nearly three years before the supernova explosion, with a relatively rapid rise in the final 100 days. The double-peaked post-explosion light curve reaches a luminosity of $\sim10^{43}~\rm erg\,s^{-1}$. The strong intermediate-width He lines observed in the nebular spectrum of SN 2023fyq imply the interaction is still active at late phases. We found that the precursor activity in SN 2023fyq is best explained by the mass transfer in a binary system involving a low-mass He star and a compact companion. An equatorial disk is likely formed in this process ($\sim$0.6$\rm M_{\odot}$), and the interaction of SN ejecta with this disk powers the main peak of the supernova. The early SN light curve reveals the presence of dense extended material ($\sim$0.3$\rm M_{\odot}$) at $\sim$3000$\rm R_{\odot}$ ejected weeks before the SN explosion, likely d
Despite their spectacular progress, language models still struggle on complex reasoning tasks, such as advanced mathematics. We consider a long-standing open problem in mathematics: discovering a Lyapunov function that ensures the global stability of a dynamical system. This problem has no known general solution, and algorithmic solvers only exist for some small polynomial systems. We propose a new method for generating synthetic training samples from random solutions, and show that sequence-to-sequence transformers trained on such datasets perform better than algorithmic solvers and humans on polynomial systems, and can discover new Lyapunov functions for non-polynomial systems.
We present an analysis of the QUARKS survey sample, focusing on protoclusters where Hot Molecular Cores (HMCs, traced by CH3CN(12--11)) and UC HII regions (traced by H30α/H40α) coexist. Using the high-resolution, high-sensitivity 1.3 mm data from the QUARKS survey, we identify 125 Hot Molecular Fragments (HMFs), which represent the substructures of HMCs at higher resolution. From line integrated intensity maps of CH3CN(12--11) and H30α, we resolve the spatial distribution of HMFs and UC HII regions. By combining with observations of CO outflows and 1.3 mm continuum, we classify HMFs into four types: HMFs associated with jet-like outflow, with wide-angle outflow, with non-detectable outflow, and shell-like HMFs near UC HII regions. This diversity possibly indicates that the hot core could be polymorphic and long-standing phenomenon in the evolution of massive protostars. The separation between HMFs and H30α/H40αemission suggests that sequential high-mass star formation within young protoclusters is not likely related to feedback mechanisms.
We consider the long-standing question of finding a parameter of a class of probability distributions that characterizes its PAC learnability. We provide a rather surprising answer - no such parameter exists. Our techniques allow us to show similar results for several general notions of characterizing learnability and for several learning tasks. We show that there is no notion of dimension that characterizes the sample complexity of learning distribution classes. We then consider the weaker requirement of only characterizing learnability (rather than the quantitative sample complexity function). We propose some natural requirements for such a characterization and go on to show that there exists no characterization of learnability that satisfies these requirements for classes of distributions. Furthermore, we show that our results hold for various other learning problems. In particular, we show that there is no notion of dimension characterizing (or characterization of learnability) for any of the tasks: classification learning for distribution classes, learning of binary classifications w.r.t. a restricted set of marginal distributions, and learnability of classes of real-valued fu
Mutation testing has been demonstrated to be one of the most powerful fault-revealing tools in the tester's tool kit. Much previous work implicitly assumed it to be sufficient to re-compute mutant suites per release. Sadly, this makes mutation results inconsistent; mutant scores from each release cannot be directly compared, making it harder to measure test improvement. Furthermore, regular code change means that a mutant suite's relevance will naturally degrade over time. We measure this degradation in relevance for 143,500 mutants in 4 non-trivial systems finding that, on overage, 52% degrade. We introduce a mutant brittleness measure and use it to audit software systems and their mutation suites. We also demonstrate how consistent-by-construction long-standing mutant suites can be identified with a 10x improvement in mutant relevance over an arbitrary test suite. Our results indicate that the research community should avoid the re-computation of mutant suites and focus, instead, on long-standing mutants, thereby improving the consistency and relevance of mutation testing.
Analytic and computational methods developed within statistical physics have found applications in numerous disciplines. In this letter, we use such methods to solve a long-standing problem in statistical genetics. The problem, posed by Haldane and Waddington [J.B.S. Haldane and C.H. Waddington, Genetics 16, 357-374 (1931)], concerns so-called recombinant inbred lines (RILs) produced by repeated inbreeding. Haldane and Waddington derived the probabilities of RILs when considering 2 and 3 genes but the case of 4 or more genes has remained elusive. Our solution uses two probabilistic frameworks relatively unknown outside of physics: Glauber's formula and self-consistent equations of the Schwinger-Dyson type. Surprisingly, this combination of statistical formalisms unveils the exact probabilities of RILs for any number of genes. Extensions of the framework may have applications in population genetics and beyond.
We solve a long-standing problem by enumerating the number of non-degenerate Desargues configurations. We extend the result to the more difficult case involving Desargues blockline structures in Section 8. A transparent proof of Desargues theorem in the plane and in space is presented as a by-product of our methods.
We propose a scenario where blazars are classified as flat-spectrum radio quasars (FSRQs), BL Lacs, low synchrotron, or high synchrotron peaked objects according to a varying mix of the Doppler boosted radiation from the jet, the emission from the accretion disk, the broad line region, and the light from the host galaxy. In this framework the peak energy of the synchrotron power (nu_peak) in blazars is independent of source type and of radio luminosity. We test this new approach, which builds upon unified schemes, using extensive Monte Carlo simulations and show that it can provide simple answers to a number of long-standing issues including, amongst others, the different cosmological evolution of BL Lacs selected in the radio and X-ray bands, the larger nu_peak values observed in BL Lacs, the fact that high synchrotron peaked blazars are always of the BL Lac type, and the existence of FSRQ/BL Lac transition objects. Objects so far classified as BL Lacs on the basis of their observed weak, or undetectable, emission lines are of two physically different classes: intrinsically weak lined objects, more common in X-ray selected samples, and heavily diluted broad lined sources, more fre
We find that significant incompleteness in stellar number counts results in a significant overestimate of the microlensing optical depth $τ$ and event rate per star per year $Γ$ toward the Galactic bulge from the first two years of the MOA-II survey. We find that the completeness in Red Clump Giant (RCG) counts $f_{\rm RC}$ decreases proportional to the galactic latitude $b$, as $f_{\rm RC}=(0.63\pm0.11)-(0.052\pm0.028)\times b$, ranging between 1 and 0.7 at $b=-6^\circ\sim-1.5^\circ$. The previous measurements using all sources by Difference Image Analysis (DIA) by MACHO and MOA-I suffer the same bias. On the other hand, the measurements using a RCG sample by OGLE-II, MACHO and EROS were free from this bias because they selected only the events associated with the resolved stars. Thus, the incompleteness both in the number of events and stellar number count cancel out. We estimate $τ$ and $Γ$ by correcting this incompleteness. In the central fields with $|l|<5^\circ$, we find $Γ=[18.74\pm0.91]\times10^{-6}\exp[(0.53\pm0.05)(3-|b|)]$ star$^{-1}$ yr$^{-1}$ and $τ_{200}=[1.84\pm0.14]\times10^{-6}\exp[(0.44\pm0.07)(3-|b|)]$ for the 427 events with $t_{\rm E}\leq200\,$days using all
A detailed analysis of the open charm effects on the decays of $J/ψ(ψ^\prime)\to VP$ is presented, where $V$ stands for light vector meson and $P$ for light pseudoscalar meson. These are the channels that the so-called "12% rule" of perturbative QCD (pQCD) is obviously violated. Nevertheless, they are also the channels that violate the pQCD helicity selection rule (HSR) at leading order. In this work, we put constraints on the electromagnetic (EM) contribution, short-distance contribution from the $c\bar{c}$ annihilation at the wavefunction origin, and long-distance contribution from the open charm threshold effects on these two decays. We show that interferences among these amplitudes, in particular, the destructive interferences between the short-distance and long-distance strong amplitudes play a key role to evade the HSR and cause the significant deviations from the pQCD expected "12% rule".
Formation energies of charged point defects in semiconductors are calculated using periodic supercells, which entail a divergence arising from long-range Coulombic interactions. The divergence is typically removed by the so-called jellium approach. Recently, Wu, Zhang and Pantelides [WZP, Phys. Rev. Lett. 119, 105501 (2017)] traced the origin of the divergence to the assumption that charged defects are formed by physically removing electrons from or adding electrons to the crystal, violating charge neutrality, a key principle of statistical mechanics that determines the Fermi level. An alternative theory was constructed by recognizing that "charged" defects form by trading carriers with the energy bands, whereby supercells are always charge-neutral so that no divergence is present and no ad-hoc procedures need to be adopted for calculations. Here we give a more detailed exposition of the foundations of both methods and show that the jellium approach can be derived from the statistical-mechanics-backed WZP definition by steps whose validity cannot be assessed a priori. In particular, the divergence appears when the charge density of band carriers is dropped, leaving a supercharged c
We model the roadway of a suspension bridge as a thin rectangular plate and we study in detail its oscillating modes. The plate is assumed to be hinged on its short edges and free on its long edges. Two different kinds of oscillating modes are found: longitudinal modes and torsional modes. Then we analyze a fourth order hyperbolic equation describing the dynamics of the bridge. In order to emphasize the structural behavior we consider an isolated equation with no forcing and damping. Due to the nonlinear behavior of the cables and hangers, a structural instability appears. With a finite dimensional approximation we prove that the system remains stable at low energies while numerical results show that for larger energies the system becomes unstable. We analyze the energy thresholds of instability and we show that the model allows to give answers to several questions left open by the Tacoma collapse in 1940.
Extending the generation horizon of video diffusion models to long sequences remains a long-standing and important challenge. Existing training-free approaches fall into two categories: extensions of bidirectional models, which are tightly coupled to specific architectures and suffer from quality degradation over long horizons, and autoregressive models, which accumulate drift errors due to exposure bias and tend to produce repetitive motion patterns. To address these issues, we propose a novel but simple inference-time approach for long video generation that is architecture-agnostic and requires no additional training. Our method generates long videos via overlapping sliding windows, where predicted clean samples from adjacent windows are blended via \emph{Tweedie matching} to enforce both \textbf{manifold constraint and temporal consistency} across overlap regions. \emph{Stochastic early-phase sampling} then synchronizes per-window trajectories by injecting fresh noise after each Tweedie matching correction in the high-noise phase, before transitioning to deterministic ODE sampling to preserve fine-grained visual fidelity. Applied to various video generation models, our method ge
Despite the growing success of Large Speech Language Models (LSLMs) in processing short-term acoustic signals, their extension to long-form audio understanding is severely bottlenecked. This limitation stems from the limited context length and the exorbitant memory footprints required for long-form inference. In this work, we propose Speech-XL, a new model that capitalizes on the intrinsic key-value (KV) sparsification capacity of Large Language Models (LLMs) to achieve high-ratio speech input compression. Specifically, we introduce a novel special token, the Speech Summarization Token (SST), for each speech interval to encapsulate the intra-interval speech information into its associated KV pairs. The SST module is trained via instruction fine-tuning, employing a curriculum learning strategy where the SST learns to compress information in a progressive manner--advancing from low-ratio (simple) to high-ratio (challenging) compression. Despite utilizing significantly less training data than other baselines, our model achieves highly competitive performance on major benchmarks, including LongSpeech and AUDIOMARATHON. By addressing the long-standing bottlenecks in long-form audio mode
Chinese stand-up comedy generation goes beyond plain text generation, requiring culturally grounded humor, precise timing, stage-performance cues, and implicit multi-step reasoning. Moreover, commonly used Chinese humor datasets are often better suited for humor understanding and evaluation than for long-form stand-up generation, making direct supervision misaligned with the target task. To address these challenges, we present OpenMic, an end-to-end multi-agent system built on AutoGen that transforms a user-provided life topic into a 3-5 minute Chinese stand-up performance and further produces a narrated comedy video. OpenMic orchestrates multiple specialized agents in a multi-round iterative loop-planning to jointly optimize humor, timing, and performability. To mitigate the dataset-task mismatch, we augment generation with retrieval-augmented generation (RAG) for material grounding and idea expansion, and we fine-tune a dedicated JokeWriter to better internalize stand-up-specific setup-punchline structures and long-range callbacks.
The shape of an object is of fundamental interest and high importance, but is not a straightforward subject if the object is on quantum scale. We here discuss how a shaped micro-object can be looked at within quantum mechanics. For this purpose, atomic nuclei are suitable, because they are tiny shaped objects. The majority of atomic nuclei are shaped like ellipsoids. Although an ellipsoid is oriented in a direction classically, such a nucleus is pointing in all directions with certain probabilities in quantum eigenstates, fulfilling rotational symmetry. This makes the direct observation of shapes formidably difficult. Here, we show, including examples, that the ellipsoidal nucleus is basically standing in a fixed direction for finite time \sim some 10^{-23} sec, as a robust consequence of time-dependent Schrodinger equation in quantum mechanics and a well-known rotational feature of nuclei. This consequence not only provides Relativistic Heavy-Ion Collisions9 with experimental feasibilities, but also leads to a deeper general understanding of stationary states with restored broken symmetry: time-dependent symmetry-breaking (e.g., ellipsoid shape) properties arise from stationary st
Self-organization through noisy interactions is ubiquitous across physics, mathematics, and machine learning, yet how long-range structure emerges from local noisy dynamics remains poorly understood. Here, we investigate three paradigmatic random-organizing particle systems drawn from distinct domains: models from soft matter physics (random organization, biased random organization) and machine learning (stochastic gradient descent), each characterized by distinct sources of noise. We discover universal long-range behavior across all systems, namely the suppression of long-range density fluctuations, governed solely by the noise correlation between particles. Furthermore, we establish a connection between the emergence of long-range order and the tendency of stochastic gradient descent to favor flat minima -- a phenomenon widely observed in machine learning. To rationalize these findings, we develop a fluctuating hydrodynamic theory that quantitatively captures all observations. Our study resolves long-standing questions about the microscopic origin of noise-induced hyperuniformity, uncovers striking parallels between stochastic gradient descent dynamics on particle system energy l
We provide solid evidence for the long-standing presumption that model Hamiltonians with short-range interactions faithfully reproduce the physics of the long-range Coulomb interaction in real materials. For this aim, we address a generic Hubbard model that captures the quantum phase transitions between metal, Mott insulator, and charge-density-wave insulator, in the absence of Fermi-surface nesting. By comparing the quantum phase diagrams for the $1/r$-Hubbard model on a half-filled chain with nearest-neighbor and $1/r$-long-range interactions, we argue that the inclusion of long-range interactions is not crucial for a proper description of interacting many-electron systems. To this end, we employ the Density Matrix Renormalization Group method on finite lattices and antiperiodic boundary conditions to determine the quantum phase transitions between the metallic Luttinger liquid for weak interactions, the Mott-Hubbard insulator for dominant on-site interactions, and the charge-density wave insulator for dominant inter-site interactions. The two phase diagrams qualitatively agree inasmuch as the quantum phase transitions are continuous in both cases. Moreover, simple Hartree-Fock t
The long-tailed problem is a long-standing challenge in Sequential Recommender Systems (SRS) in which the problem exists in terms of both users and items. While many existing studies address the long-tailed problem in SRS, they only focus on either the user or item perspective. However, we discover that the long-tailed user and item problems exist at the same time, and considering only either one of them leads to sub-optimal performance of the other one. In this paper, we propose a novel framework for SRS, called Mutual Enhancement of Long-Tailed user and item (MELT), that jointly alleviates the long-tailed problem in the perspectives of both users and items. MELT consists of bilateral branches each of which is responsible for long-tailed users and items, respectively, and the branches are trained to mutually enhance each other, which is trained effectively by a curriculum learning-based training. MELT is model-agnostic in that it can be seamlessly integrated with existing SRS models. Extensive experiments on eight datasets demonstrate the benefit of alleviating the long-tailed problems in terms of both users and items even without sacrificing the performance of head users and item