arXiv subjects

Yueying Li

Publications and source records attributed to Yueying Li.

At least 19 recordsLinked to original sources

arXiv2026

PowerSlider: Exploiting Phase Asymmetry for LLM Serving under Demand Response

AI inference clusters are increasingly constrained by instantaneous power, not just energy: grid operators condition new capacity on demand response, imposing time-varying power caps. Existing LLM serving systems optimize a static energy objective or shed fixed priority tiers under load; either way, goodput collapses when the power envelope moves. An LLM pipeline is not a uniform load: compute-bound prefill loses throughput almost linearly with GPU frequency, memory-bound answer decode sustains it down to $0.57\times$ nominal, and reasoning's thinking phase couples KV-cache capacity to scheduling -- so a cap should be steered to where each watt costs the least performance. PowerSlider does so with a new Flex SLO contract that turns bounded user slack into an optimization constraint, prefill--think--answer disaggregation exposing per-stage frequency and KV control, and a Karush--Kuhn--Tucker (KKT) online solver re-solving within 7.7 ms of every cap change, backed by a consolidated fail-safe that power-gates drained instances when DVFS bottoms out on static power. On SGLang with production traces, \sys{} sustains 78.3\% online goodput at a 30\% cap reduction versus 47.6\% for the best of five baselines ($1.64\times$), holds latency-critical tails within $1.3\times$ of nominal (baselines: $2.3$--$6\times$, up to $12\times$), and delivers 92\% mean goodput through a replayed CAISO grid-emergency day bottoming at $0.41\times$ (54\% at the trough; every baseline below 7\%).

cs.DC↗

arXiv2026

Beyond Zooming: Learning Multi-Tool Visual Reasoning for Ultra-High-Resolution Remote Sensing

Ultra-high-resolution (UHR) remote-sensing (RS) imagery provides fine-grained Earth-observation evidence over city-scale scenes, but poses a fundamental challenge for multimodal large language models (MLLMs): task-relevant evidence is often sparse, local, and spatially dispersed across extremely large visual contexts. A natural solution is to equip MLLMs with zoom-in tools for active local inspection. However, through a pilot study on XLRS-Bench, we find that zoom-in is only partially effective: it resolves easy and medium-level tasks with locally recoverable evidence, but saturates on hard cases requiring global search, multi-region comparison, path planning, or dispersed-evidence reasoning. Motivated by this finding, we move beyond single-tool zoom-in and introduce GeoMTVR, a large-scale Geospatial Multi-Tool Visual Reasoning dataset built from wide-area satellite imagery. GeoMTVR contains 13K UHR VQA samples with interleaved reasoning trajectories, diverse visual tool calls, and returned visual observations, enabling models to learn question decomposition, tool selection, regional inspection, object-level grounding, auxiliary visual reasoning, and cross-tool evidence integration. Beyond supervised fine-tuning, we propose a tool-attention-focused reinforcement learning algorithm that concentrates optimization on critical tool-use decisions, including when to invoke tools, which tool to select, where to apply it, and how to interpret tool outputs. By combining SFT on GeoMTVR with our RL algorithm, we develop GeoLens, a multi-tool visual reasoning MLLM for UHR RS. Experiments show that GeoLens consistently outperforms direct reasoning and single-tool zoom-in baselines, achieving stronger accuracy, better evidence grounding, and more efficient tool-use trajectories.

cs.CV↗

arXiv2026

Obshazard-bench: Benchmarking Multimodal Foundation Models for Real-Time Disaster Intelligence from Raw Earth Observation Streams

Multimodal Large Language Models (MLLMs) are increasingly used to interpret Earth observation data, yet their capability to support real-world disaster emergency response remains insufficiently evaluated. Existing remote sensing benchmarks largely rely on static, post-hoc, and expert-processed products, such as gridded reanalysis data, which are difficult to align with operational disaster scenarios where hazards evolve rapidly and decisions must be made under strict time constraints. To bridge this gap, we introduce Obshazard-bench, a real-time, observation-driven benchmark for evaluating disaster intelligence in MLLMs. Unlike image-centric or post-event benchmarks, Obshazard-bench directly integrates raw, high-frequency satellite sounding streams from diverse satellite sensors with concurrent ground-station observations, historical disaster records, and socio-economic indicators, bypassing delayed expert-processing and physical-inversion pipelines. The benchmark covers 8 major disaster categories and 28 sub-categories across more than 60 countries, incorporating over 120 historically documented extreme-event cases and thousands of lifecycle-oriented VQA samples. Moreover, Obshazard-bench further defines a three-stage evaluation taxonomy aligned with the operational disaster workflow: Predictive Crisis Anticipation for pre-disaster risk detection and early forecasting, Active Evolution Reasoning for in-situ disaster tracking and termination prediction, and Multi-faceted Impact Quantification for post-disaster magnitude deduction, humanitarian burden estimation, and socio-economic impact assessment. Experiments on representative general-purpose and Earth-focused foundation models reveal substantial limitations in transforming raw multi-channel physical observations into temporally grounded and decision-relevant disaster reasoning.

cs.CL↗

arXiv2026

Beyond Prediction: Tail-Aware Scheduling for LLM Inference

LLM serving exhibits extreme length variability, making size-based scheduling difficult in practice. Recent LLM schedulers approximate SJF/SRPT using predicted decode lengths or ranks and primarily report mean-centric metrics such as TTFT and TBT. We show that these prediction-driven policies can be fragile under distribution shifts, bursty arrivals, and GPU memory pressure, while offering limited control over the tail latency (P90-P99) that dominates user experience, even with perfect decode-length knowledge. We introduce a distribution-aware, prediction-free scheduling framework that replaces explicit length prediction with soft priority boosting driven by lightweight statistical signals. Our design co-optimizes scheduling and cache-aware preemption to account for memory-coupled decode dynamics across workload mixes. Evaluated on production and open-source traces, our method reduces P99 TTLT by up to 35-50% relative to SRPT with perfect length knowledge and reduces TTFT by 34-47% across workloads, including reasoning-heavy and chat-heavy tasks. These results demonstrate a robust alternative for optimizing tail latency in online LLM serving.

cs.LG↗

arXiv2026

Bipolar-doped superconducting infinite-layer cuprates

Distilling the intrinsic physics of the superconducting CuO2 plane from the complexities of charge-reservoir layers is a defining challenge in high-temperature superconductivity. While superconducting electron-doped infinite-layer cuprates have been synthesized, controllable and uniform hole doping has long remained elusive despite exploratory attempts, limiting spectroscopic insights. Here, we realize bipolar doping across infinite-layer (Sr,Eu)CuO2 and (Ca,Li)CuO2+δ single-crystalline thin films, mapping the electronic phase diagram. Both electron- and hole-doped films show pronounced electrical resistance anisotropy, indicating the quasi-two-dimensional nature of the CuO2 planes. Angle-resolved photoemission spectroscopy across electron- and hole-doped regimes reveals persistent antiferromagnetic band folding coexisting with superconductivity. Remarkably, at a hole doping ~0.07 determined by Luttinger volume, the antiferromagnetic folding emerges from Fermi arcs within the film's single Fermi surface, with the onset superconducting transition temperature exceeding 60 K. These findings redefine the interplay between magnetic order and superconductivity and establish a definitive platform to investigate the intrinsic mechanism of high-temperature superconducting cuprates.

cond-mat.supr-con↗

arXiv2026

SVSR: A Self-Verification and Self-Rectification Paradigm for Multimodal Reasoning

Current multimodal models often suffer from shallow reasoning, leading to errors caused by incomplete or inconsistent thought processes. To address this limitation, we propose Self-Verification and Self-Rectification (SVSR), a unified framework that explicitly integrates self-verification and self-rectification into the model's reasoning pipeline, substantially improving robustness and reliability in complex visual understanding and multimodal reasoning tasks. SVSR is built on a novel three-stage training paradigm. First, we construct a high-quality unified preference dataset by refining reasoning traces from pre-trained vision-language models, incorporating both forward and backward reasoning to embed self-reflective signals. Second, we perform cold-start supervised fine-tuning on this dataset to learn structured, multi-step reasoning behaviors. Third, we apply a Semi-online Direct Preference Optimization (Semi-online DPO) process, continuously augmenting the training corpus with high-quality, model-generated reasoning traces filtered by a powerful teacher VLM. This pipeline enables the model to learn, elicit, and refine its ability to self-verify and self-rectify. Extensive experiments across diverse benchmarks demonstrate that SVSR improves reasoning accuracy and enables stronger generalization to unseen tasks and question types. Notably, once trained with explicit self-reflective reasoning, the model also exhibits improved implicit reasoning ability, outperforming strong baselines even when no explicit reasoning traces are provided. These results highlight the potential of SVSR for building more dependable, introspective, and cognitively aligned multimodal systems.

cs.AI↗

arXiv2026

Throughput-Optimal Scheduling Algorithms for LLM Inference and AI Agents

As demand for Large Language Models (LLMs) and AI agents grows rapidly, optimizing systems for efficient LLM inference becomes critical. While significant efforts have targeted system-level engineering, little has been explored from a mathematical modeling and queueing perspective. In this paper, we develop the queueing fundamentals for LLM inference. In particular, we study the throughput aspect of LLM inference systems. We prove that a large class of `work-conserving' scheduling algorithms achieve maximum throughput for both individual requests and AI-agent workloads with directed acyclic graph (DAG) and fork-join routing topologies, establishing `work-conserving' as a key design principle for practitioners. Technically, we develop a fluid-limit framework for multi-class batched processing networks under $K$-FCFS scheduling, which may be of independent interest. Evaluations of real-world systems confirm that Orca and Sarathi-Serve are throughput-optimal, reassuring practitioners, while FasterTransformer and vanilla vLLM are not maximally stable and should be used with caution. Our analysis also reveals how constraints such as batch size limits and cyclic routing topologies complicate the throughput picture, pointing to rich open questions at the intersection of queueing theory and LLM system design.

stat.ML↗

arXiv2026

Electronic structures across superconductor-insulator transition in Ruddlesden-Popper bilayer nickelate films

High-transition-temperature ($T_{C}$) superconductivity is recently discovered in Ruddlesden-Popper (RP) nickelate films with extraordinarily strong oxidation. While investigating phase diagrams is essential for uncovering the superconducting mechanism, the oxygen-tuned superconductor-insulator transition (SIT) in RP nickelates differs fundamentally from that in cuprates or iron-based systems. Here, we unveil the evolution of electronic structure in RP bilayer nickelate thin films across the SIT, combining angle-resolved photoemission spectroscopy (ARPES) and X-ray absorption spectroscopy (XAS) for both occupied and unoccupied states. In the superconducting state, a coherent quasiparticle band near Fermi level ($E_{F}$) coexists with an incoherent waterfall feature at high energy, paralleling that in cuprates. Approaching the insulating state with oxygen deficiency, the spectral weight of the occupied coherent quasiparticle band is gradually suppressed, accompanied by pronounced density of states redistribution and orbital reconfiguration in unoccupied states. These results reveal the electronic origin of the SIT in the phase diagram, which transcends carrier doping effects and oxygen vacancy states. Our findings point to a decisive role of oxygen in shaping the essential electronic landscape of RP bilayer nickelates, offering crucial insights into the superconducting mechanism.

cond-mat.supr-con↗

arXiv2026

Semantic-Geometric Dual Compression: Training-Free Visual Token Reduction for Ultra-High-Resolution Remote Sensing Understanding

Multimodal Large Language Models (MLLMs) have demonstrated immense potential in Earth observation. However, the massive visual tokens generated when processing Ultra-High-Resolution (UHR) imagery introduce prohibitive computational overhead, severely bottlenecking their inference efficiency. Existing visual token compression methods predominantly adopt static and uniform compression strategies, neglecting the inherent "Semantic-Geometric Duality" in remote sensing interpretation tasks. Specifically, object semantic tasks focus on the abstract semantics of objects and benefit from aggressive background pruning, whereas scene geometric tasks critically rely on the integrity of spatial topology. To address this challenge, we propose DualComp, a task-adaptive dual-stream token compression framework. Dynamically guided by a lightweight pre-trained router, DualComp decouples feature processing into two dedicated pathways. In the object semantic stream, the Spatially-Contiguous Semantic Aggregator (SCSA) utilizes size-adaptive clustering to aggregates redundant background while protecting small object. In the scene geometric stream, the Instruction-Guided Structure Recoverer (IGSR) introduces a greedy path-tracing topology completion mechanism to reconstruct spatial skeletons. Experiments on the UHR remote sensing benchmark XLRS-Bench demonstrate that DualComp accomplishes high-fidelity remote sensing interpretation at an exceptionally low computational cost, achieving simultaneous improvements in both efficiency and accuracy.

cs.CV↗

arXiv2026

Superconductivity and Electronic Structures of Nickelate Thin Film Superstructures

Ruddlesden-Popper (RP) nickelates have emerged as a crucial platform for exploring the mechanisms of high-temperature superconductivity. However, the Fermi surface topology required for superconductivity remains elusive. Here, beyond the superconducting pure bilayer (2222) phase, we report the thin film growth and ambient-pressure superconductivity of monolayer-bilayer (1212) and bilayer-trilayer (2323) superstructures, together with the absence of superconductivity in monolayer-trilayer (1313) superstructure, under identical compressive epitaxial strain. The onset superconducting transition temperatures range from 46 to 50 K, exceeding the McMillan limit. Angle-resolved photoemission spectroscopy reveals key Fermi surface differences in these atomically-engineered structures. In superconducting 1212 and 2222 films, a dispersive hole-like band ($γ^{\mathrm{II}}$) forms an underlying Fermi pocket, surrounding the Brillouin zone corner. In contrast, the top of the flat band ($γ^{\mathrm{III}}$) is observed ~70 meV below $E_\text{F}$ in the non-superconducting 1313 films. Particularly, the superconducting 2323 films host both $γ^{\mathrm{II}}$ and $γ^{\mathrm{III}}$ bands. The polarization dependence of the $γ$ bands reveals their Ni $d_{z^2}$ origin. Our findings expand the family of ambient-pressure nickelate superconductors and establish a connection between structural configuration, electronic structure, and the emergence of superconductivity in nickelates.

cond-mat.supr-con↗

arXiv2026

Three-Dimensional Electronic Structures in Superconducting Ruddlesden-Popper Bilayer Nickelate Films

Beyond the quasi-two-dimensional (2D) paradigm of cuprates, the role of the third dimension of the Ruddlesden-Popper bilayer nickelates is essential to decoding their superconducting mechanism. Here, using angle-resolved photoemission spectroscopy (ARPES) with varied photon energies, we systematically investigate the electronic band structures in three dimensions for superconducting (La,Pr,Sm)$_3$Ni$_2$O$_7$/SrLaAlO$_4$ thin films (superconducting onset temperature $T_c^{\text{onset}} \sim 48$ K) transferred via a cryogenic ultra-high vacuum suitcase. We reveal an orbital-dependent dimensionality: while the $d{x^2-y^2}$-dominant bands exhibit a quasi-2D character, the $d{z^2}$-dominant band displays a finite $k_z$ dispersion. Finite energy gaps are identified on all observed bands across multiple high-symmetry directions. Systematic temperature-dependent analysis characterizes the superconducting nature of the gap on the $d{z^2}$-derived band, revealing a large gap $Δ\sim 18$ meV and a ratio $2Δ/k_BT_c\sim 8$ exceeding the weak-coupling BCS limit. The suppression of spectral weight near the Fermi level persists above the superconducting transition temperature. Ubiquitous waterfall-like spectral features evidence the presence of electron interactions. These results underscore the role of the $d_{z^2}$ orbital and correlations, placing constraints on theoretical models for nickelate superconductivity.

cond-mat.supr-con↗

arXiv2026

M3D-BFS: a Multi-stage Dynamic Fusion Strategy for Sample-Adaptive Multi-Modal Brain Network Analysis

Multi-modal fusion is of great significance in neuroscience which integrates information from different modalities and can achieve better performance than uni-modal methods in downstream tasks. Current multi-modal fusion methods in brain networks, which mainly focus on structural connectivity (SC) and functional connectivity (FC) modalities, are static in nature. They feed different samples into the same model with identical computation, ignoring inherent difference between input samples. This lack of sample adaptation hinders model's further performance. To this end, we innovatively propose a multi-stage dynamic fusion strategy (M3D-BFS) for sample-adaptive multi-modal brain network analysis. Unlike other static fusion methods, we design different mixture-of-experts (MoEs) for uni- and multi-modal representations where modules can adaptively change as input sample changes during inference. To alleviate issue of MoE where training of experts may be collapsed, we divide our method into 3 stages. We first train uni-modal encoders respectively, then pretrain single experts of MoEs before finally finetuning the whole model. A multi-modal disentanglement loss is designed to enhance the final representations. To the best of our knowledge, this is the first work for dynamic fusion for multi-modal brain network analysis. Extensive experiments on different real-world datasets demonstrates the superiority of M3D-BFS.

cs.AI↗

arXiv2026

Scaling Reward Modeling without Human Supervision

Learning from feedback is an instrumental process for advancing the capabilities and safety of frontier models, yet its effectiveness is often constrained by cost and scalability. We present a pilot study that explores scaling reward models through unsupervised approaches. We operationalize reward-based scaling (RBS), in its simplest form, as preference learning over document prefixes and suffixes drawn from large-scale web corpora. Its advantage is demonstrated in various aspects: despite using no human annotations, training on 11M tokens of math-focused web data yields steady gains on RewardBench v1 and v2, and these improvements consistently transfer across diverse initialization backbones spanning model families and scales. Across models, our method improves RewardBench v2 accuracy by up to +7.7 points on average, with gains of up to +16.1 on in-domain math subsets and consistent improvements on out-of-domain safety and general subsets. When applied to best-of-N selection and policy optimization, these reward models substantially improve downstream math performance and match or exceed strong supervised reward model baselines of similar size. Overall, we demonstrate the feasibility and promise of training reward models without costly and potentially unreliable human annotations.

cs.LG↗

arXiv2026

GeoEyes: On-Demand Visual Focusing for Evidence-Grounded Understanding of Ultra-High-Resolution Remote Sensing Imagery

The "thinking-with-images" paradigm enables multimodal large language models (MLLMs) to actively explore visual scenes via zoom-in tools. This is essential for ultra-high-resolution (UHR) remote sensing VQA, where task-relevant cues are sparse and tiny. However, we observe a consistent failure mode in existing zoom-enabled MLLMs: Tool Usage Homogenization, where tool calls collapse into task-agnostic patterns, limiting effective evidence acquisition. To address this, we propose GeoEyes, a staged training framework consisting of (1) a cold-start SFT dataset, UHR Chain-of-Zoom (UHR-CoZ), which covers diverse zooming regimes, and (2) an agentic reinforcement learning method, AdaZoom-GRPO, that explicitly rewards evidence gain and answer improvement during zoom interactions. The resulting model learns on-demand zooming with proper stopping behavior and achieves substantial improvements on UHR remote sensing benchmarks, with 54.23% accuracy on XLRS-Bench.

cs.CV↗

arXiv2026

OmniEarth-Bench: Towards Holistic Evaluation of Earth's Six Spheres and Cross-Spheres Interactions with Multimodal Observational Earth Data

Existing benchmarks for multimodal learning in Earth science offer limited, siloed coverage of Earth's spheres and their cross-sphere interactions, typically restricting evaluation to the human-activity sphere of atmosphere and to at most 16 tasks. These limitations: narrow-source heterogeneity (single/few data sources), constrained scientific granularity, and limited-sphere extensibility. Therefore, we introduce OmniEarth-Bench, the first multimodal benchmark that systematically spans all six spheres: atmosphere, lithosphere, oceanosphere, cryosphere, biosphere, and human-activity sphere, and cross-spheres. Built with a scalable, modular-topology data inference framework and native multi-observation sources and expert-in-the-loop curation, OmniEarth-Bench produces 29,855 standardized, expert-curated annotations. All annotations are organized into a four-level hierarchy (Sphere, Scenario, Ability, Task), encompassing 109 expert-curated evaluation tasks. Experiments on 9 state-of-the-art MLLMs reveal that even the most advanced models struggle with our benchmarks, where none of them reach 35% accuracy, revealing systematic gaps in Earth-system cognitive ability. The dataset and evaluation code were released at OmniEarth-Bench (https://anonymous.4open.science/r/OmniEarth-Bench-B1BD).

cs.CV↗

arXiv2026

Text Before Vision: Staged Knowledge Injection Matters for Agentic RLVR in Ultra-High-Resolution Remote Sensing Understanding

Multimodal reasoning for ultra-high-resolution (UHR) remote sensing (RS) is usually bottlenecked by visual evidence acquisition: the model necessitates localizing tiny task-relevant regions in massive pixel spaces. While Agentic Reinforcement Learning with Verifiable Rewards (RLVR) using zoom-in tools offers a path forward, we find that standard reinforcement learning struggles to navigate these vast visual spaces without structured domain priors. In this paper, we investigate the interplay between post-training paradigms: comparing Cold-start Supervised Fine-Tuning (SFT), RLVR, and Agentic RLVR on the UHR RS benchmark.Our controlled studies yield a counter-intuitive finding: high-quality Earth-science text-only QA is a primary driver of UHR visual reasoning gains. Despite lacking images, domain-specific text injects the concepts, mechanistic explanations, and decision rules necessary to guide visual evidence retrieval.Based on this, we propose a staged knowledge injection recipe: (1) cold-starting with scalable, knowledge-graph-verified Earth-science text QA to instill reasoning structures;and (2) "pre-warming" on the same hard UHR image-text examples during SFT to stabilize and amplify subsequent tool-based RL. This approach achieves a 60.40% Pass@1 on XLRS-Bench, significantly outperforming larger general purpose models (e.g., GPT-5.2, Gemini 3.0 Pro, Intern-S1) and establishing a new state-of-the-art.

cs.AI↗

arXiv2026

Superconductor-insulator transitions in infinite-layer nickelates controlled via ${operando}$ monitored reduction

Nickelates represent an emerging class of superconductors that demand innovative approaches for structural and electronic phase modulations. Continuous control over superconductor-insulator transition (SIT) in nickelates remains particularly challenging, hindering both fundamental understanding and potential applications. Here, we demonstrate SIT in infinite-layer nickelate superconductors utilizing multiple techniques, including an ${operando}$ monitored reduction (OMR) method. OMR enables ultrawide-range continuous modulation of the Ni 3${d}$ orbital electron occupancy from ~3${d}^7$ to ~3${d}^9$. The 3${d}$ occupancy is calibrated through systematic synchrotron X-ray absorption (XAS), combined with scanning transmission electron microscopy (STEM) annular bright field (ABF) analysis of oxygen atoms. SIT is further modulated via ionic liquid gating and magnetic field. Strikingly different from cuprates, our Nernst effect measurements show that pairing initiates at the onset of the resistive drop. The subsequent emergence of the Meissner effect at zero resistance marks the establishment of global phase coherence. Angle-dependent magnetotransport within the transition temperature regime indicates a mixture of two-dimensional (2D) and three-dimensional (3D) superconducting characters, suggesting the observed SIT deviates from the canonical 2D model. Our results provide a unique perspective on the interplay of structural and electronic phase transitions in the infinite-layer nickelates across the oxygen content-magnetic field-temperature parameter space.

cond-mat.supr-con↗

arXiv2025

GeoLLaVA-8K: Scaling Remote-Sensing Multimodal Large Language Models to 8K Resolution

Ultra-high-resolution (UHR) remote sensing (RS) imagery offers valuable data for Earth observation but pose challenges for existing multimodal foundation models due to two key bottlenecks: (1) limited availability of UHR training data, and (2) token explosion caused by the large image size. To address data scarcity, we introduce SuperRS-VQA (avg. 8,376$\times$8,376) and HighRS-VQA (avg. 2,000$\times$1,912), the highest-resolution vision-language datasets in RS to date, covering 22 real-world dialogue tasks. To mitigate token explosion, our pilot studies reveal significant redundancy in RS images: crucial information is concentrated in a small subset of object-centric tokens, while pruning background tokens (e.g., ocean or forest) can even improve performance. Motivated by these findings, we propose two strategies: Background Token Pruning and Anchored Token Selection, to reduce the memory footprint while preserving key semantics.Integrating these techniques, we introduce GeoLLaVA-8K, the first RS-focused multimodal large language model capable of handling inputs up to 8K$\times$8K resolution, built on the LLaVA framework. Trained on SuperRS-VQA and HighRS-VQA, GeoLLaVA-8K sets a new state-of-the-art on the XLRS-Bench.

cs.CV↗