Qikai Zhang | Searcharxiv

arXiv2026

Efficient Evaluation of Gravitational Lensing Amplification Factors: A Deep Learning Framework

Wave optics is essential for analyzing lensed gravitational waves (GWs), yet evaluating the diffraction integral $F(ω, y)$ is computationally expensive. We present a Sinusoidal Representation Networks (SIRENs) framework for the dimensionless amplification factor, demonstrating its efficacy and generalization through Point Mass Lens (PML) and Singular Isothermal Sphere (SIS) test cases. Unlike standard architectures that suffer from spectral bias, the network's periodic activation functions structurally align with the integral's oscillatory kernel, effectively resolving high-frequency spectral features. The resulting estimator achieves $\mathcal{O}(10^{-3})$ relative accuracy and a $\sim 100\times$ speedup compared to direct numerical integration. By shifting the computational burden to offline training, our framework yields a stable $\mathcal{O}(1)$ inference complexity. This guarantees constant, sub-millisecond evaluation times even in the weak-lensing diffraction tail where traditional methods stagnate. Additionally, the dimensionless formulation ensures intrinsic scale invariance, enabling direct application across astrophysical regimes from stellar-mass lenses in the ground-based LVK band to supermassive black holes in the space-based LISA band.

astro-ph.IM↗

arXiv2026

Time-Domain Deep Learning for Pairwise Identification of Strongly Lensed Gravitational-Wave Candidates

As gravitational wave (GW) catalogs continue to expand, exhaustive Bayesian comparisons of candidate event pairs become increasingly computationally expensive, which motivates the development of fast prescreening methods for strongly lensed GW searches. We formulate lensed-pair identification as a binary verification problem using two preprocessed strain segments. To address this task, we propose Physics-Inspired ResNet (PI-ResNet), a Siamese one-dimensional residual network for pairwise GW candidate classification. Unlike spectrogram-based prescreening approaches, PI-ResNet operates directly on whitened time-domain strain data and avoids an intermediate time--frequency image representation. A shared residual backbone with Squeeze-and-Excitation (SE) modules encodes the two input segments, and the paired embeddings are compared through absolute feature differences and Hadamard-product interactions. We train and evaluate the model using simulated GW signals from binary black hole mergers lensed by point-mass (PM) and singular isothermal sphere (SIS) lenses, injected into simulated LIGO and Einstein Telescope (ET) detector noise. Under ET design noise, PI-ResNet achieves accuracies of $95.60\%$ for SIS lenses and $93.80\%$ for PM lenses, while maintaining $84.03\%$ and $78.25\%$ accuracy under simulated LIGO H1--L1 Gaussian noise. These results suggest that direct learning from 1D strain data provides an efficient and physically motivated preselection statistic for candidate lensed GW pairs, while also indicating the need for detector-domain adaptation.

astro-ph.HE↗