Searcharxiv⌕ Search

arXiv subjects

Song Li

Publications and source records attributed to Song Li.

At least 19 recordsLinked to original sources
arXiv2026

I$_4$ complex acting as near-infrared emitter with a quantum memory in silicon

Silicon color centers emitting in the near infrared offer a route to spin--photon interfaces compatible with mature complementary metal--oxide--semiconductor fabrication. A key unresolved candidate is the X-line at 1039.7 meV in the photoluminescence (PL) spectrum, long attributed to a complex of four self-interstitials (I$_4$), yet whose microscopic structure and spin properties have remained controversial and uncharacterized at a predictive level. Using hybrid-functional (HSE06) density functional theory, we unambiguously establish the bare I$_4$ complex as the origin of the X-line: our calculations quantitatively reproduce the zero-phonon line, the complete phonon sideband structure---identifying the prominent secondary peaks as one- and two-phonon replicas of strongly coupled quasi-local modes---and the EPR hyperfine constants. The charge transition level is found at VBM$+$0.08~eV, conclusively ruling out the previously proposed association with the B3 EPR center and its correlated deep donor level. Most significantly, we identify a metastable spin-1 triplet excited state separated from the singlet by $\sim$1.6~meV, consistent with the NL51 EPR signal. The $D_{2d}$ symmetry of the defect suppresses intersystem crossing and phosphorescence by first-order spin--orbit coupling, implying an exceptionally long spin--lattice relaxation time. Combined with a spectrally resolvable zero-field splitting and near-infrared emission, the triplet state of the I$_4$ complex constitutes a viable spin quantum memory intrinsic to silicon.

cond-mat.mtrl-sci↗
arXiv2026

RESPLE2: Inertial-Informed Recursive Non-Uniform Spline Estimation for Multi-Sensor Odometry with Global Refinement

We present RESPLE2, a recursive multi-sensor state estimation framework using non-uniform B-splines for 6-DoF motion representation. By exploiting the knot-clamping boundary property, we propose a theoretically grounded mechanism for exact control-point prediction directly informed by inertial kinematics along the trajectory, with knot placement adapted to motion and measurement availability. Within an iterated extended Kalman filter, we perform this prediction using a subset of inertial measurements. A subsequent reset step preserves trajectory consistency as the knots extend. The update step further incorporates heterogeneous sensor measurements at their native timestamps through continuous-time observation models, supporting flexible combinations of LiDARs, cameras, inertial measurement units, and leg kinematics. We further develop a lightweight global optimization module that directly refines non-uniform B-spline control points by incorporating loop-closure constraints derived from LiDAR-visual cues. We conduct extensive evaluations on public and experimental datasets across handheld, tracked, legged, and hybrid aerial-legged platforms, covering highly dynamic maneuvers with extreme accelerations. Compared with state-of-the-art multi-sensor odometry systems, RESPLE2 achieves competitive or superior estimation accuracy and robustness in the corresponding sensor configurations, particularly under aggressive motion, while maintaining real-time operation. We publicly release our code and experimental datasets at https://github.com/ASIG-X/RESPLE2.

cs.RO↗
arXiv2026

A hybrid quantum interferometer bridging Hong-Ou-Mandel and N00N-state interference

Hong-Ou-Mandel (HOM) interference offers robustness in high-noise environments due to its phase-insensitive nature, while N00N-state interference provides enhanced phase sensitivity for precision measurements in low-noise settings. However, their realizations have typically required separate experimental platforms. Here, we report a single hybrid quantum interferometer that can dynamically transition between HOM and N00N-state interference by adjusting two time delays, $τ_1$ and $τ_2$. This tunability enables operation in phase-insensitive, phase-sensitive, or hybrid modes within a unified setup, as confirmed by theoretical analysis and experimental demonstration. Our system offers a versatile platform with potential for quantum sensing applications, paving the way for integrated quantum technologies capable of adapting to diverse noise conditions.

quant-ph↗
arXiv2026

A Blind Trust, the Bloody Thrust: When Attacker-Controlled Hook Updates Steer AI Agent Harnesses towards Malicious Behaviors

Modern AI agent harnesses expose lifecycle hooks that bind shell commands to runtime events such as session start, tool calls, and file edits. These commands run with host privileges yet ship as lifecycle-hook configuration and may fire at times the LLM never observes. We identify the lifecycle-hook update path, which harnesses trust blindly, as a new attack surface. Under a supply-chain threat model in which an attacker controls only plugin metadata and lifecycle-hook configuration, a benign versioned plugin can be trojanized by an update that silently binds attacker-chosen commands to benign events, yielding malicious host-side behavior such as privilege escalation. We propose HookPry, an open-source and fully automated attack framework that systematically exploits this vulnerability across heterogeneous AI agent harnesses. HookPry realizes ten attack objectives; across 25 combinations of harnesses and backends in 1,000 end-to-end runs, it compromises all seven evaluated harnesses, with per-harness success rates reaching 92.5%. Representative defenses remain insufficient: Microsoft Defender has 0% recall, and the union of three static defenses misses 47.5% of malicious artifacts.

cs.CR↗
arXiv2026

Likelihood-Constrained Acoustic Reranking for Training-Free Hallucination Mitigation in LLM-Based ASR

Large language model (LLM)-based automatic speech recognition (ASR) systems achieve strong performance on conventional speech data by leveraging powerful linguistic priors and multilingual capabilities. However, under challenging conditions, these priors can override acoustic evidence, resulting in unintended translation, instruction execution, repetition, or catastrophic deletion. We propose Likelihood-Constrained Acoustic Reranking (LCAR), a training-free decoding method that improves acoustic grounding while preserving support from the base model. At each decoding step, LCAR first retains tokens whose base-model likelihood falls within a margin of the greedy token, then reranks them using an acoustic compatibility score computed from attention-pooled audio embeddings and the existing LM head. By restricting acoustic intervention to plausible, model-supported alternatives, LCAR requires no additional training, external detector, reference transcript, or auxiliary model at inference. We evaluate LCAR on four LLM-based ASR systems using human-audited TTS and open-source speech challenge suites. At $δ=0.60$, LCAR removes 38.8--57.1\% of detector-identified hallucination failures while largely maintaining WER/CER on standard open-source test sets.

eess.AS↗
arXiv2026

VIPER-MCP: Detecting and Exploiting Taint-Style Vulnerabilities in Model Context Protocol Servers

Model Context Protocol (MCP) has emerged as a standard interface for connecting LLM agents to external tools. Because MCP servers expose privileged operations such as shell execution, network access, and file-system manipulation to agent-driven invocation, implementation flaws in tool handlers can create a direct path from natural-language input to security-sensitive sinks, potentially granting attackers remote code execution or full system compromise. Existing approaches either produce unconfirmed static alerts without dynamic validation, or rely on fixed template libraries that lack code-level guidance and fail to trigger vulnerabilities requiring specific parameter shapes or multi-step taint paths. In this paper, we present VIPER-MCP, the first end-to-end automated vulnerability auditing framework for MCP servers that not only detects taint-style vulnerabilities but also dynamically confirms their exploitability by producing concrete proof-of-concept prompts. VIPER-MCP introduces two novel techniques: (1) an anchor-query pass in a two-pass static analysis strategy that augments standard taint alerts with function-level structural context, resolving file-level static artifacts to specific MCP tool handlers and producing vulnerability-anchored call chains; and (2) a feedback-driven prompt evolution mechanism that employs dual-mutator scheduling that independently corrects tool-selection drift and deepens parameter penetration, together with fitness-scored seed selection to iteratively refine natural-language prompts toward vulnerable sinks. In a large-scale scan of 39,884 real-world open-source MCP server repositories, VIPER-MCP discovered 106 0-day vulnerabilities, all of which were confirmed through end-to-end exploit traces, with 67 CVE IDs assigned to date.

cs.CR↗
arXiv2026

PhaseLift for Coded Diffraction Patterns: Optimal Sampling Rate

Recovering a complex-valued signal from coded diffraction patterns, namely the Fourier intensities obtained after modulating the signal with a collection of masks, is a fundamental structured phase retrieval problem arising in diffraction imaging and related applications. Despite its practical importance, the theoretical analysis of this structured framework remains scarce. In the standard random mask model, the optimal sampling rate achievable by computationally tractable recovery methods has remained open. In this paper, we establish the optimal sampling rate for the PhaseLift feasibility program. More precisely, PhaseLift achieves exact recovery of an unknown signal $\pmb{x}_0\in\mathbb{C}^n$, up to a global phase, from $\mathcal{O}(\log n)$ random masks, with polynomially decaying failure probability. Since $Ω(\log n)$ masks are necessary to identify certain signals under the erasure mask ensemble, our result thereby achieves the optimal mask complexity. Equivalently, PhaseLift attains the optimal total sampling rate of $m=\mathcal{O}( n\log n)$ scalar intensity measurements. The proof is based on an approximate dual certificate construction via a refined golfing scheme that combines adaptive mask allocation with a dimension-independent truncation threshold.

cs.IT↗
arXiv2026

Low-Rank Matrix Recovery via Heavy-Tailed Quadratic Sampling

The problem of recovering an (approximately) low-rank Hermitian matrix $\pmb{M}_0 \in \mathbb{C}^{n \times n}$ of rank $r$ from quadratic sampling matrices of the form $\{\pmb{a}_k \pmb{a}_k^*\}_{k=1}^m$ arises in a variety of applications, including phase retrieval. To obtain rigorous recovery guarantees, the sampling vectors $\{\pmb{a}_k\}_{k=1}^m$ are typically modeled probabilistically. However, most existing theoretical results rely on Gaussian or sub-Gaussian assumptions, which may not accurately capture practical data models. In many applications, sampling vectors exhibit heavier tails, while theoretical understanding in such regimes remains scarce. In this paper, we bridge this gap. We show that two widely used convex approaches, nuclear norm minimization and semidefinite-constrained empirical risk minimization, achieve uniform, stable, and robust recovery under the mild assumption that the entries of the sampling vectors have only finite $4+δ$ moments, with the optimal sample complexity $m = \mathcal{O}(rn)$ up to moment-dependent constants. The two main ingredients of our analysis are moment estimates for quadratic forms established via decoupling, together with recent advances in covariance estimation in heavy-tailed settings. As byproducts, we also establish the optimal sample complexity for low-rank matrix recovery under complex projective $4$-design sampling, thereby improving upon previous results, and obtain stability guarantees for phase retrieval under similarly weak moment assumptions.

math.ST↗
arXiv2026

Measurability of Quadrupole Deviations from Kerr in Binary black hole Mergers

We investigate the measurability of black hole quadrupole deviations from Kerr using five binary black hole mergers observed by the LIGO-Virgo-KAGRA Collaboration with the beyond-general-relativity full-waveform model $Ψ_{\mathrm{FD}}$. While earlier lower-SNR events mildly favored nonzero quadrupole deviations, the newly included high-SNR GWTC-4 events GW231226, GW230814, and GW250114 yield results increasingly consistent with the Kerr prediction. In particular, GW230814 and GW250114, the two highest-SNR events in our sample, yield deviations consistent with zero. We further perform separate inspiral and post-inspiral analyses and find both the posterior distributions centered close to $ΔQ/Q=0$ for GW230814 and GW250114. Overall, the full-waveform, inspiral, and post-inspiral results for GW230814 and GW250114 reveal no observable departure from the no-hair theorem within the sensitivity of the current data and the $Ψ_{\mathrm{FD}}$ framework. Although the limited number of events prevents a definitive conclusion, future detections of additional high-SNR binary black hole mergers will enable increasingly stringent and robust tests of the Kerr nature of black holes.

gr-qc↗
arXiv2026

Can a secluded self-interacting dark sector generate detectable gravitational waves?

In this work we study the possibility to detect the gravitational waves generated by a secluded self-interacting dark sector. ``Secluded'' means that the dark sector has almost no portal to the visible sector and thus its entropy is conserved by itself, and ``self-interacting'' means that dark matter in this model has a significant interaction to itself, making it consistent with the small-scale structure observations. A spontaneously broken $U(1)'$ is introduced for the interactions in the dark sector, and nearly massless dark radiation is also introduced to avoid the over-closure problem. Through a parameter space scan, we find that this model is highly constrained by the currently observed effective number of neutrinos ($N_{\text{eff}}$) and the large-scale structure observable Lyman-$α$. Together, these two constraints exclude a large parameter space that is favored by future gravitational-wave detectors, but there is still a small portion of the model parameter space that can be detected by SKA.

hep-ph↗
arXiv2026

Dimensionality-Driven Charge Stabilization of Group-IV Color Centers in Diamond Ultrathin Films

Neutral group-IV vacancy (XV, X = Si, Ge, Sn, and Pb) centers in diamond are emerging solid-state spin-photon interfaces because of their favorable spin coherence and inversion symmetry-protected optical transitions. However, stabilizing their neutral charge state typically requires stringent Fermi-level engineering in high purity boron-doped diamond, which poses significant materials-growth challenges. Here, we demonstrate that dimensional confinement in diamane provides an alternative route to charge-state stabilization without intentional doping. Using first-principles calculations, we show that quantum confinement and surface termination cooperatively tune the host band gap and shift the occupied defect states upward from the valence-band edge, thereby enlarging the thermodynamic stability window of the neutral charge state and suppressing valence-band assisted excitation pathways. We further reveal that the thickness and surface termination of diamane enable systematic tuning of the electronic structure, zero-field splitting, and spin-orbit coupling of XV centers while largely preserving their optical transition energies. Among the structures considered, hydrogenated diamane offers the most favorable balance between charge-state stability and magneto-optical performance. More broadly, our findings establish dimensional confinement as a general strategy for engineering the charge, optical, and spin properties of solid-state quantum defects.

cond-mat.mtrl-sci↗
arXiv2026

Convergence of Spectral Descent for Non-smooth Optimization

The Muon optimizer has recently demonstrated remarkable empirical success in training large language models. However, the theoretical understanding of its mechanisms remains limited. Current convergence guarantees for Muon rely heavily on smoothness assumptions, leaving its non-smooth convergence behavior largely unexplored. In this work, we take a step toward bridging this gap by investigating Spectral Descent (SD), a simplified variant of Muon, together with its truncated counterpart, Truncated Spectral Descent (TSD). Under convexity, Lipschitz continuity, and sharpness conditions, we establish global linear convergence for both SD and TSD in non-smooth convex formulations. We also study regularized variants equipped with decoupled weight decay and derive sublinear convergence guarantees through their connection with Frank-Wolfe methods. Finally, we apply our theoretical framework to robust low-rank matrix recovery under mixed sparse and dense noise regimes and provide rigorous recovery guarantees. Numerical experiments support the theoretical findings and demonstrate the effectiveness of Muon-type methods for non-smooth optimization.

cs.LG↗
arXiv2026

Interacting donor-acceptor pairs as the origin of coupled spin-optical signals in hexagonal boron nitride

Optically addressable spin defects in hexagonal boron nitride hold promise for room-temperature quantum technologies, but their microscopic identities remain largely unknown. Using first principles calculations, we show that coupled spin optical signals arise from interacting donor acceptor pairs, not the commonly believed isolated defects. Intra and inter pair separations control charge transfer, electronic structure, and spin coupling, thereby greatly modulating zero phonon lines, phonon sidebands, lifetimes, and the sign of optically detected magnetic resonance contrast. Importantly, we identify two distinct charge-state-dependent coupling regimes and extend this picture to correlated defect ensembles, explaining the wide diversity of experimental observations. Our results establish a microscopic framework for coupled defect behavior and provide design principles for spin-active quantum emitters in wide bandgap semiconductors.

cond-mat.mtrl-sci↗
arXiv2026

Low-Rank Toeplitz Matrix Restoration: Descent Cone Analysis and Structured Random Matrix

This note demonstrates that we can stably recover all symmetric Toeplitz matrices $\pmb{X}_0\in\mathbb{R}^{n\times n}$ of rank at most $r$ from a number of rank-one subgaussian measurements on the order of $r\log^{2} n$ with an exponentially decreasing failure probability by employing a nuclear norm minimization program. Our approach utilizes descent cone analysis through Mendelson's small ball method with the Toeplitz constraint. The key ingredient is to determine the spectral norm of a random matrix with Toeplitz structure, which may be of independent interest. This improves upon earlier analyses and resolves the conjecture in Chen et al. (IEEE Transactions on Information Theory, 61(7):4034--4059, 2015).

cs.IT↗
arXiv2026

PlantMarkerBench: A Multi-Species Benchmark for Evidence-Grounded Plant Marker Reasoning

Cell-type-specific marker genes are fundamental to plant biology, yet existing resources primarily rely on curated databases or high-throughput studies without explicitly modeling the supporting evidence found in scientific literature. We introduce PlantMarkerBench, a multi-species benchmark for evaluating literature-grounded plant marker evidence interpretation from full-text biological papers. PlantMarkerBench is constructed using a modular curation pipeline integrating large-scale literature retrieval, hybrid search, species-aware biological grounding, structured evidence extraction, and targeted human review. The benchmark spans four plant species -- Arabidopsis, maize, rice, and tomato -- and contains 5,550 sentence-level evidence instances annotated for marker-evidence validity, evidence type, and support strength. We define two benchmark tasks: determining whether a candidate sentence provides valid marker evidence for a gene-cell-type pair, and classifying the evidence into expression, localization, function, indirect, or negative categories. We benchmark diverse open-weight and closed-source language models across species and prompting strategies. Although frontier models achieve relatively strong performance on direct expression evidence, performance drops substantially on functional, indirect, and weak-support evidence, with evidence-type confusion emerging as a dominant failure mode. Open-weight models additionally exhibit elevated false-positive rates under ambiguous biological contexts. PlantMarkerBench provides a challenging and reproducible evaluation framework for literature-grounded biological evidence attribution and supports future research on trustworthy scientific information extraction and AI-assisted plant biology.

cs.CL↗
arXiv2026

Geometric Characteristics and Stable Guarantees for Phaseless Operators and Structured Matrix Restoration

In this paper, we first propose a unified framework for analyzing the stability of the phaseless operators for both amplitude and intensity measurement on an arbitrary geometric set, thereby characterizing the robust performance of phase retrieval via the empirical minimization method. We introduce the random embedding of concave lifting operators to characterize the unified analysis of any geometric set. Similarly, we investigate the robust performance of structured matrix restoration problem through the robust injectivity of a linear rank one measurement operator on an arbitrary matrix set. The core of our analysis is to establish unified empirical chaos processes characterization for various matrix sets. Talagrand's $γ_α$-functionals are employed to characterize the connection between the geometric constraints and the number of measurements required for stability or robust injectivity. We also construct adversarial noise to demonstrate the sharpness of the recovery bounds derived through the empirical minimization method in the both scenarios.

cs.IT↗
arXiv2026

How Label Imbalance Shapes Geometry: A General Spectral Analysis of Multi-Label Neural Collapse

This work investigates the phenomenon of Neural Collapse (NC) in multi-label classification, extending its conceptual framework from multi-class learning to general correlated and imbalanced multi-label settings. Although recent studies have identified a ''tag-wise averaging'' structure for multi-label features, this view relies on implicit assumptions of label balance and combinatorial symmetry. Consequently, it fails to account for the geometrical distortions caused by intrinsic label correlations and data imbalance, which are common in practice. We resolve the multiplicity-one imbalance conjecture raised by Li et al. (2024), showing that higher-multiplicity prototypes obey a class-frequency-weighted synthesis rule rather than uniform averaging. To address this, we propose a rigorous spectral-control framework to analyze the terminal phase of multi-label learning under general imbalanced conditions. We introduce the label covariance spectrum $κ_m$, a scalar controlling the distribution-dependent lower-bound geometry, derived from the second-order moment matrix of the label distribution. Contrary to the averaging perspective, our analysis reveals that the centered label covariance spectrum controls the stability of terminal geometry by quantifying the weakest centered inter-class contrast directions. We prove that the classical Tag-wise Averaging emerges only as a special case under perfect orthogonality. Numerical experiments on synthetic distributions validate our theoretical bounds. This work resolves the scaled-average aspect of the imbalance conjecture and establishes a unifying theoretical framework that extends Neural Collapse to complex, imbalanced multi-label settings.

cs.LG↗
arXiv2026

SmartGuard: Leveraging Large Language Models for Network Attack Detection through Audit Log Analysis and Summarization

End-point monitoring solutions are widely deployed in today's enterprise environments to support advanced attack detection and investigation. These monitors continuously record system-level activities as audit logs and provide deep visibility into security events. Unfortunately, existing methods of semantic analysis based on audit logs have low granularity, only reaching the system call level, making it difficult to effectively classify highly covert behaviors. Additionally, existing works mainly match audit log streams with rule knowledge bases describing behaviors, which heavily rely on expertise and lack the ability to detect unknown attacks and provide interpretive descriptions. In this paper, we propose SmartGuard, an automated method that combines abstracted behaviors from audit event semantics with large language models. SmartGuard extracts specific behaviors (function level) from incoming system logs and constructs a knowledge graph, divides events by threads, and combines event summaries with graph embeddings to achieve information diagnosis and provide explanatory narratives through large language models. Our evaluation shows that SmartGuard achieves an average F1 score of 96\% in assessing malicious behaviors and demonstrates good scalability across multiple models and unknown attacks. It also possesses excellent fine-tuning capabilities, allowing experts to assist in timely system updates.

cs.CR↗