J. Zhu

X. Feng, J. Zhu, M. Wu et al.

Without incurring significant computational overhead, train-free long video generation aims to enable foundation video generation models to produce longer videos. Frame-level autoregressive frameworks, e.g., FIFO-diffusion, offer the advantage of generating infinitely long videos with constant memory consumption. However, the mismatch between training and inference, coupled with the challenge of maintaining long-term consistency, limits the effective utilization of foundation models. To mitigate these concerns, we propose \textbf{MIGA}, a novel infinite-frame long video generation method. Firstly, we propose an effective two-stage alignment mechanism that mitigates the training-inference gap by reducing the excessive noise span fed to the model. We then introduce an innovative dual consistency enhancement mechanism, where the self-reflection approach corrects early high-noise frames and the long-range frame guidance approach leverages later low-noise frames with broad coverage to steer generation, jointly improving temporal consistency. Extensive experiments on VBench and NarrLV demonstrate the state-of-the-art performance of MIGA. Our project page is available at https://xiaokunfeng.github.io/miga_homepage/.

T. Schibler, J. Xue, J. Zhu

Recently, motivated by the rapid increase of the data size in various applications, Monemizadeh [APPROX'23] and Driemel, Monemizadeh, Oh, Staals, and Woodruff [SoCG'25] studied geometric problems in the setting where the only access to the input point set is via querying a range-search oracle. Algorithms in this setting are evaluated on two criteria: (i) the number of queries to the oracle and (ii) the error of the output. In this paper, we continue this line of research and investigate one of the most fundamental geometric problems in the oracle setting, i.e., the convex hull problem. Let $P$ be an unknown set of points in $[0,1]^d$ equipped with a range-emptiness oracle. Via querying the oracle, the algorithm is supposed to output a convex polygon $C \subseteq [0,1]^d$ as an estimation of the convex hull $CH(P)$ of $P$. The error of the output is defined as the volume of the symmetric difference $C \oplus CH(P) = (C \backslash CH(P)) \cup (CH(P) \backslash C)$. We prove tight and near-tight tradeoffs between the number of queries and the error of the output for different variants of the problem, depending on the type of the range-emptiness queries and whether the queries are non-adaptive or adaptive. - Orthogonal emptiness queries in $d$-dimensional space: We show that the minimum error a deterministic algorithm can achieve with $q$ queries is $Θ(q^{-1/d})$ if the queries are non-adaptive, and $Θ(q^{-1/(d-1)})$ if the queries are adaptive. In particular, in 2D, the bounds are $Θ(1/\sqrt{q})$ and $Θ(1/q)$ for non-adaptive and adaptive queries, respectively. - Halfplane emptiness queries in 2D: We show that the minimum error a deterministic algorithm can achieve with $q$ queries is $Θ(1/\sqrt{q})$ if the queries are non-adaptive, and $\widetildeΘ(1/q^2)$ if the queries are adaptive. Here $\widetildeΘ(\cdot)$ hides logarithmic factors.

S. A. K. Leeney, H. T. J. Bevins, E. de Lera Acedo et al. · oxford

Radiometers are crucial instruments in radio astronomy, forming the primary component of nearly all radio telescopes. They measure the intensity of electromagnetic radiation, converting this radiation into electrical signals. A radiometer's primary components are an antenna and a Low Noise Amplifier (LNA), which is the core of the ``receiver'' chain. Instrumental effects introduced by the receiver are typically corrected or removed during calibration. However, impedance mismatches between the antenna and receiver can introduce unwanted signal reflections and distortions. Traditional calibration methods, such as Dicke switching, alternate the receiver input between the antenna and a well-characterised reference source to mitigate errors by comparison. Recent advances in Machine Learning (ML) offer promising alternatives. Neural networks, which are trained using known signal sources, provide a powerful means to model and calibrate complex systems where traditional analytical approaches struggle. These methods are especially relevant for detecting the faint sky-averaged 21-cm signal from atomic hydrogen at high redshifts. This is one of the main challenges in observational Cosmology today. Here, for the first time, we introduce and test a machine learning-based calibration framework capable of achieving the precision required for radiometric experiments aiming to detect the 21-cm line.

J. Zhu, N. M. Lockmann, M. K. Czwalinna et al.

Longitudinal properties of electron bunches are critical for the performance of a wide range of scientific facilities. In a free-electron laser, for example, the existing diagnostics only provide very limited longitudinal information of the electron bunch during online tuning and optimization. We leverage the power of artificial intelligence to build a neural network model using experimental data, in order to bring the destructive longitudinal phase space (LPS) diagnostics online virtually and improve the existing current profile online diagnostics which uses a coherent transition radiation (CTR) spectrometer. The model can also serve as a digital twin of the real machine on which algorithms can be tested efficiently and effectively. We demonstrate at the FLASH facility that the encoder-decoder model with more than one decoder can make highly accurate predictions of megapixel LPS images and coherent transition radiation spectra concurrently for electron bunches in a bunch train with broad ranges of LPS shapes and peak currents, which are obtained by scanning all the major control knobs for LPS manipulation. Furthermore, we propose a way to significantly improve the CTR spectrometer online measurement by combining the predicted and measured spectra. Our work showcases how to combine virtual and real diagnostics in order to provide heterogeneous and reliable mixed diagnostics for scientific facilities.

J. Zhu, C. Rea, R. S. Granetz et al.

Next generation high performance (HP) tokamaks risk damage from unmitigated disruptions at high current and power. Achieving reliable disruption prediction for a device's HP operation based on its low performance (LP) data is key to success. In this letter, through explorative data analysis and dedicated numerical experiments on multiple existing tokamaks, we demonstrate how the operational regimes of tokamaks can affect the power of a trained disruption predictor. First, our results suggest data-driven disruption predictors trained on abundant LP discharges work poorly on the HP regime of the same tokamak, which is a consequence of the distinct distributions of the tightly correlated signals related to disruptions in these two regimes. Second, we find that matching operational parameters among tokamaks strongly improves cross-machine accuracy which implies our model learns from the underlying scalings of dimensionless physics parameters like q_{95}, β_{p} and confirms the importance of these parameters in disruption physics and cross machine domain matching from the data-driven perspective. Finally, our results show how in the absence of HP data from the target devices, the best predictivity of the HP regime for the target machine can be achieved by combining LP data from the target with HP data from other machines. These results provide a possible disruption predictor development strategy for next generation tokamaks, such as ITER and SPARC, and highlight the importance of developing on existing machines baseline scenario discharges of future tokamaks to collect more relevant disruptive data.