Yutong Geng | Searcharxiv

arXiv2026

Remote dispersion scan: transformer-network retrieval of ultrafast pulses after non-linear propagation

Accurate and rapid characterization of broadband electric fields is essential for all ultrafast applications and remains an active field of research. In this work, we introduce remote dispersion scan, a transformer neural network enabled dispersion scan based pulse characterization method that can characterize femtosecond laser pulses. A local scan of the non-linear spectral phase before several linear and nonlinear processes, including amplification, compression, and self phase modulation, allows for the field retrieval remotely at the interaction region. We show that the reconstruction accuracy obtained from a single measurement of the fundamental and second harmonic is comparable to that of a full two dimensional scan. We confirm the technique experimentally by compressing a 300 W, 1.3 ps, 1030 nm pulse in a hollow core fiber to 100 fs and measuring the fundamental and second harmonic spectra while scanning the second order phase in a pulse shaper before power amplification. These results establish a simple, robust, alignment insensitive live-view pulse reconstruction modality.

physics.optics↗

arXiv2025

Online Flow Time Minimization: Tight Bounds for Non-Preemptive Algorithms

This paper studies the online scheduling problem of minimizing total flow time for $n$ jobs on $m$ identical machines. A classical $\Omega(n)$ lower bound shows that no deterministic single-machine algorithm can beat the trivial greedy, even when $n$ is known in advance. However, this barrier is specific to deterministic algorithms on a single machine, leaving open what randomization, multiple machines, or the kill-and-restart capability can achieve. We give a nearly complete answer. For randomized non-preemptive algorithms, we establish a tight $\Theta(\sqrt{n/m})$ competitive ratio, which also improves the best offline approximation to $O(\sqrt{n/m})$. For deterministic non-preemptive algorithms on multiple machines, we prove an $O(n/m^2 + \sqrt{n/m}\log m)$ upper bound and an $\Omega(n/m^2 + \sqrt{n/m})$ lower bound. In the kill-and-restart model, we reveal a sharp transition for deterministic algorithms: $\Omega(n/\log n)$ for $m = 1$ versus $\Theta(\sqrt{n/m})$ for $m \ge 2$; the latter matches the optimal randomized ratio, and we further show that randomization provides no additional power in this model. We also investigate the setting where $n$ is unknown. We prove that no randomized non-preemptive algorithm achieves $o(n)$ on one machine or $o(n/m^2 + \sqrt{n/m})$ on $m$ machines. In contrast, our kill-and-restart algorithm achieves $O(n^{\alpha}/\sqrt{m})$ for $m \ge 2$, where $\alpha = (\sqrt{5}-1)/2$, breaking the trivial bound without knowledge of $n$.

cs.DS↗