arXiv subjects

Siying Li

Publications and source records attributed to Siying Li.

At least 19 recordsLinked to original sources

arXiv2026

SenseNova-U1.5: Towards Native Unified Visual Intelligence

We launch SenseNova-U1.5, an 8B-MoT native unified multimodal model that understands, reasons about, and generates visual content within an encoder-free and VAE-free architecture. We strengthen its visual interface through spatially coherent patch reconstruction and scale its training with carefully curated generation and editing data, improved task formulation, structural prompt enhancement, and native resolutions of up to 4K. For post-training, we optimize specialized experts for visual aesthetics, bilingual text rendering, infographic generation, and image editing, and consolidate their capabilities through multi-expert on-policy distillation. Across extensive evaluations, SenseNova-U1.5 largely advances image fidelity, text rendering, complex composition, multi-reference editing, and interleaved generation, while improving instruction following and preserving subject identity, geometry, and unmodified regions. Despite limited exposure to structured formats in its generation data, SenseNova-U1.5 generalizes effectively to long, complex, and structured visual instructions, further proving that multimodal understanding can transfer to visual planning and creation. Together, these findings position native unified modelling as a promising path towards systems that perceive, reason and create within a fully end-to-end framework. We will open-source training code, including supervised fine-tuning, reinforcement learning, and on-policy distillation.

cs.CV↗

arXiv2026

Separating Stream Stability from Long-Term Recall in Language Models

Methods for streaming language models are often discussed alongside long-context and memory systems, although they solve different problems. An attention sink can stabilize autoregressive generation over an indefinitely long stream while the model remains unable to use content that has left its recent-token cache. We argue that this distinction should be explicit in system claims and evaluation. We introduce three horizons: the stability horizon, over which predictive behavior remains well behaved; the access horizon, over which past content can still causally affect the output; and the utility horizon, over which a task retains acceptable performance. We show constructively that the stability horizon can be infinite while the access and utility horizons are finite. We then propose ThreeH, an evaluation contract that measures all three horizons under a common state and compute budget. Applying the framework to attention-sink streaming clarifies its strength, constant-memory, stable generation, without treating anchor tokens as semantic memory. The framework exposes roles for cache policies, recurrent state, retrieval, and external memory. Experiments on 128K-token streams, delayed binding recall, and delayed decisions show that attention sinks preserve local modeling but not content beyond the active cache; recurrent and retrieval state extend the semantic horizon.

cs.CL↗

arXiv2026

DecoupleGS: Interactive 3D Gaussian Splatting for End-to-End Autonomous Driving Testing

End-to-end (E2E) autonomous driving algorithms require rigorous closed-loop validation in simulation environments offering high visual fidelity, strong interactivity, and real-time performance. Existing approaches, from game engines to static neural rendering, inherently trade off these requirements and struggle with the dynamic scene composition essential for E2E testing. To bridge this gap, we propose a novel decoupled 3D Gaussian Splatting (3DGS) framework tailored for large-scale E2E evaluation. We fundamentally decompose scenes into a high-fidelity static background and manipulable dynamic agents using an object-centric canonical representation. To resolve resulting representational conflicts, we introduce three targeted modules: (1) asset compression via perceptual pruning and vector quantization for real-time traffic rendering; (2) map-guided geometric registration leveraging semantic topology to strictly align trajectories; and (3) proxy-based relighting transferring ambient illumination for seamless photometric integration. Extensive experiments demonstrate that DecoupleGS achieves a balanced fidelity-efficiency trade-off, improves metric and photometric consistency, and provides a practical closed-loop sensor simulation platform for E2E autonomous driving evaluation.

cs.CV↗

arXiv2026

Collision and coalescence dynamics of bosonic quantum Hall droplets

Recently bosonic quantum Hall droplets have been observed in rapidly rotating two-dimensional Bose-Einstein condensates (BECs), which exhibit robust dynamical stability. Inspired by this, we systematically investigate the collision and coalescence dynamics of these droplets within the Gross-Pitaevskii framework. For two-droplet collisions, we find two distinct collision outcomes, namely merging and separation, that are controlled by the initial relative velocity. The critical velocity exhibits a universal scaling law with the interaction and the particle number as $v_c \propto (gN)^{1/4}$, which can be interpreted from a simplified analytical model, revealing the essential role of the collision time. It differs fundamentally from the mechanism governing the conventional Lee-Huang-Yang stabilized quantum droplets. Furthermore, while the collision can change the shape of the droplet significantly, the center of mass trajectory remains nearly unaffected, owing to the conservation of angular momentum. For overlapping stationary droplets, vortex arrays can emerge through Kelvin-Helmholtz instability driven by phase-induced shear flow. Although two droplets may merge into a larger one, extended states cannot be constructed from multiple overlapping droplets. Instead, the system dynamically reorganizes into new isolated droplets, revealing the localized property in the bulk region. Our results reveal the unique nonequilibrium dynamics of quantum Hall droplets and suggest new pathways for manipulating strongly correlated rotating quantum fluids.

cond-mat.quant-gas↗

arXiv2026

SAGE: An LLM-driven Self Reflective Agentic Framework for Fraud Detection

Fraud detection in payment, e-commerce, and telecommunications systems requires accuracy at the individual level, robustness under severe class imbalance, and ease of understanding for risk managers. Existing methods fall at least one of these requirements: automated machine learning systems search a fixed numerical space without semantic awareness of the dataset; graph neural network-based methods require pre-defined relational graphs and remain opaque at the individual-decision level; and the design of general-purpose large language model (LLM) agents does not consider the recall and precision constraints specific to real-world fraud detection. In this paper, we propose SAGE, the first end-to-end LLM-driven multi-agent framework for fraud detection. SAGE coordinates three dedicated agents that make decisions based on a six-layer Data Diagnostic Tree (DDT) and a Markov decision process guided by natural-language gradients, automatically optimizing the model under a fraud-specific reward. On five fraud datasets and five LLM backbones, SAGE wins $96.00\%$ of method--dataset comparisons and improves F1 by an average of $40.86\%$ over baselines. The code is available at https://github.com/yichenC1c/SAGE.

cs.AI↗

arXiv2026

Stepwise Reasoning Enhancement for LLMs via External Subgraph Generation

Large language models have shown strong performance in natural language generation and downstream reasoning tasks, but they still struggle with logical consistency, factual grounding, and interpretability in complex multi-step reasoning. To address these limitations, this paper proposes SGR, a stepwise reasoning enhancement framework that integrates large language models with external knowledge graphs through query-relevant subgraph generation. Given an input question, SGR first extracts key entities, relations, and constraints to construct a structured schema, then retrieves compact subgraphs from a knowledge graph using schema-guided querying. The generated subgraphs provide explicit relational evidence that guides the language model through step-by-step reasoning. In addition, SGR combines direct Cypher-based reasoning with collaborative reasoning integration, allowing candidate answers from multiple reasoning paths to be validated and aggregated according to both model confidence and graph consistency. Experiments on benchmark datasets including CWQ, WebQSP, GrailQA, and KQA Pro demonstrate that SGR improves reasoning accuracy and Hits@1 performance over standard prompting and several knowledge-enhanced baselines. Ablation studies further show that schema guidance and Neo4j-based retrieval are both crucial to the effectiveness of the framework. These results indicate that dynamically generated external subgraphs can improve the accuracy, robustness, and interpretability of LLM-based reasoning.

cs.CL↗

arXiv2026

SGR: A Stepwise Reasoning Framework for LLMs with External Subgraph Generation

Large Language Models (LLMs) have demonstrated strong capabilities across diverse NLP applications, such as translation, text generation, and question answering. Nevertheless, they remain limited in complex settings that demand deep reasoning and logical inference. Since these models are trained on large-scale text corpora, their generation process may still introduce irrelevant, noisy, or factually inconsistent content. To mitigate this problem, we introduce SGR, a stepwise framework that enhances LLM reasoning through external subgraph generation. SGR builds query-specific subgraphs from external knowledge bases and uses their semantic structure to support multi-step inference. By grounding intermediate reasoning steps in structured external knowledge, the framework helps the model concentrate on relevant entities, relations, and supporting evidence. In particular, SGR first constructs a subgraph tailored to the input question. It then guides the model to reason progressively over the generated structure and combines multiple reasoning trajectories to obtain the final prediction. Experimental results across several benchmark datasets show that SGR achieves consistent improvements over competitive baselines, highlighting its value for improving both reasoning accuracy and factual reliability.

cs.CL↗

arXiv2026

SenseNova-U1: Unifying Multimodal Understanding and Generation with NEO-unify Architecture

Recent large vision-language models (VLMs) remain fundamentally constrained by a persistent dichotomy: understanding and generation are treated as distinct problems, leading to fragmented architectures, cascaded pipelines, and misaligned representation spaces. We argue that this divide is not merely an engineering artifact, but a structural limitation that hinders the emergence of native multimodal intelligence. Hence, we introduce SenseNova-U1, a native unified multimodal paradigm built upon NEO-unify, in which understanding and generation evolve as synergistic views of a single underlying process. We launch two native unified variants, SenseNova-U1-8B-MoT and SenseNova-U1-A3B-MoT, built on dense (8B) and mixture-of-experts (30B-A3B) understanding baselines, respectively. Designed from first principles, they rival top-tier understanding-only VLMs across text understanding, vision-language perception, knowledge reasoning, agentic decision-making, and spatial intelligence. Meanwhile, they deliver strong semantic consistency and visual fidelity, excelling in conventional or knowledge-intensive any-to-image (X2I) synthesis, complex text-rich infographic generation, and interleaved vision-language generation, with or without think patterns. Beyond performance, we show detailed model design, data preprocessing, pre-/post-training, and inference strategies to support community research. Last but not least, preliminary evidence demonstrates that our models extend beyond perception and generation, performing strongly in vision-language-action (VLA) and world model (WM) scenarios. This points toward a broader roadmap where models do not translate between modalities, but think and act across them in a native manner. Multimodal AI is no longer about connecting separate systems, but about building a unified one and trusting the necessary capabilities to emerge from within.

cs.CV↗

arXiv2025

A Stepwise-Enhanced Reasoning Framework for Large Language Models Based on External Subgraph Generation

Large Language Models (LLMs) have achieved strong performance across a wide range of natural language processing tasks in recent years, including machine translation, text generation, and question answering. As their applications extend to increasingly complex scenarios, however, LLMs continue to face challenges in tasks that require deep reasoning and logical inference. In particular, models trained on large scale textual corpora may incorporate noisy or irrelevant information during generation, which can lead to incorrect predictions or outputs that are inconsistent with factual knowledge. To address this limitation, we propose a stepwise reasoning enhancement framework for LLMs based on external subgraph generation, termed SGR. The proposed framework dynamically constructs query relevant subgraphs from external knowledge bases and leverages their semantic structure to guide the reasoning process. By performing reasoning in a step by step manner over structured subgraphs, SGR reduces the influence of noisy information and improves reasoning accuracy. Specifically, the framework first generates an external subgraph tailored to the input query, then guides the model to conduct multi step reasoning grounded in the subgraph, and finally integrates multiple reasoning paths to produce the final answer. Experimental results on multiple benchmark datasets demonstrate that SGR consistently outperforms strong baselines, indicating its effectiveness in enhancing the reasoning capabilities of LLMs.

cs.CL↗

arXiv2025

Risk-Based Capacity Accreditation of Resource-Colocated Large Loads in Capacity Markets

We study capacity accreditation of resource-colocated large loads, defined as large demands such as data center and manufacturing loads colocated with behind-the-meter generation and storage resources, synchronously connected to the bulk power system, and capable of participating in the wholesale electricity market as an integrated unit. Because the accredited capacity of a resource portfolio is not equal to the sum of its individual resources' capacity values, we adopt a risk-based capacity accreditation framework to evaluate the combined reliability contribution of colocated resources. Grounded in the effective load carrying capability (ELCC) metric, the proposed capacity accreditation employs a convex optimization engine that jointly dispatches colocated resources to minimize reliability risk. We apply the developed methodology to a hydrogen manufacturing facility with colocated renewable generation, storage, and fuel cell resources.

eess.SY↗

arXiv2025

Energy Management for Renewable-Colocated Artificial Intelligence Data Centers

We develop an energy management system (EMS) for artificial intelligence (AI) data centers with colocated renewable generation. Under a cost-minimizing framework, the EMS of renewable-colocated data center (RCDC) co-optimizes AI workload scheduling, on-site renewable utilization, and electricity market participation. Within both wholesale and retail market participation models, the economic benefit of the RCDC operation is maximized. Empirical evaluations using real-world traces of electricity prices, data center power consumption, and renewable generation demonstrate significant electricity cost reduction from renewable and AI data center colocations.

math.OC↗

arXiv2025

Bosonic quantum Hall droplets in rapidly rotating two-dimensional Bose-Einstein condensates

Recent experiments demonstrate that rapidly rotating Bose-Einstein condensates (BECs) near the lowest Landau level can self-organize into interaction-driven persistent quantum Hall droplet arrays. Inspired by this discovery, we investigate the formation and dynamics of single quantum Hall droplet and droplet arrays in rapidly rotating BECs. Guided by a rigorous theorem on localized many-body states for two-dimensional interacting systems in a magnetic field, we construct single quantum Hall droplet and droplet array states which are shown to be stationary solutions to the Gross-Pitaevskii equation in the rotating frame. The single quantum Hall droplet is shown to be dynamically stable, which underpins its role as the basic unit in a droplet array. The stability of the quantum Hall droplet arrays is demonstrated by their dynamic formation from a phase engineered initial condensate. Our study sheds light onto the nature of the quantum Hall droplet state in a rapidly rotating BEC and offers a new approach for generating and manipulating quantum Hall droplet arrays through designing the initial condensate phase.

cond-mat.quant-gas↗

arXiv2025

Renewable-Colocated Green Hydrogen Production: Optimal Scheduling and Profitability

We study the optimal green hydrogen production and energy market participation of a renewable-colocated hydrogen producer (RCHP) that utilizes onsite renewable generation for both hydrogen production and grid services. Under deterministic and stochastic profit-maximization frameworks, we analyze RCHP's multiple market participation models and derive closed-form optimal scheduling policies that dynamically allocate renewable energy to hydrogen production and electricity export to the wholesale market. Analytical characterizations of the RCHP's operating profit and the optimal sizing of renewable and electrolyzer capacities are obtained. We use real-time renewable generation and electricity price data from three independent system operators to evaluate the impacts of market prices and environmental policies on RCHP's profitability.

eess.SY↗

arXiv2025

Are There High-Density Deep States in AtomicLayer-Deposited IGZO Thin Film?

It has been well recognized that there exist high-density deep states in IGZO thin films. Many of the device characteristics of IGZO transistors, such as negative bias illumination stability (NBIS),were understood to be related to these deep states. However, in this work, it is found that deep state density (NtD) of atomic-layer-deposited (ALD) IGZO transistors can be an ultra-low value (2.3*10^12 /cm^3) by the proposed NBIS-free light assisted I-V measurements so that the deep states do not affect the I-V characteristics even in subthreshold region. This work also reveals that NBIS is not related to the photoexcitation of electrons in deep states. Our results suggest that the existence of deep states and the impact of deep states on ALD IGZO transistors may need to be revisited.

cond-mat.mtrl-sci↗

arXiv2024

Decentralized Equitable Energy Access in Energy Communities

We address the issue of equitable energy access within an energy community consisting of members with diverse socioeconomic backgrounds, including varying income levels and differing capacities to access distributed energy resources such as solar power and storage systems. While optimal energy consumption scheduling is well-studied, integrating equity into decentralized real-time energy access remains under-explored. This paper formulates Equity-regarding Welfare Maximization (EqWM)--a welfare optimization energy scheduling subject to equity constraints. We further develop a decentralized implementation (D-EqWM) as a bi-level optimization, where a non-profit operator designs a community pricing policy aimed at maximizing overall welfare, subject to constraints that ensure equitable access. Community members, in turn, optimize their individual consumption based on these prices. We present the optimal pricing policy along with its key properties.

eess.SY↗

arXiv2024

Optimal regularity of subsonic steady-states solution of Euler-Poisson equations for semiconductors with sonic boundary

In this paper, we study the optimal regularity of the stationary sonic-subsonic solution to the unipolar isothermal hydrodynamic model of semiconductors with sonic boundary. Applying the comparison principle and the energy estimate, we obtain the regularity of the sonic-subsonic solution as $C^{\frac{1}{2}}[0,1]\cap W^{1,p}(0,1)$ for any $p<2$, which is then proved to be optimal by analyzing the property of solution around the singular point on the sonic line, i.e., $ρ\notin C^ν[0,1]$ for any $ν>\frac{1}{2}$, and $ρ\notin W^{1,κ}(0,1)$ for any $κ\ge 2$. Furthermore, we explore the influence of the semiconductors effect on the singularity of solution at sonic points $x=1$ and $x=0$, that is, the solution always has strong singularity at sonic point $x=1$ for any relaxation time $τ>0$, but, once the relaxation time is sufficiently large $τ\gg 1$, then the sonic-subsonic steady-states possess the strong singularity at both sonic boundaries $x=0$ and $x=1$. We also show that the pure subsonic solution $ρ$ belongs to $W^{2,\infty}(0,1)$, which can be embedded into $C^{1,1}[0,1]$, and it is much better than the regularity of sonic-subsonic solutions.

math.AP↗

arXiv2024

Multi-Interval Energy-Reserve Co-Optimization with SoC-Dependent Bids from Battery Storage

We consider the problem of co-optimized energy-reserve market clearing with state-of-charge (SoC) dependent bids from battery storage participants. While SoC-dependent bids capture storage's degradation and opportunity costs, such bids result in a non-convex optimization in the market clearing process. More challenging is the regulation reserve capacity clearing, where the SoC-dependent cost is uncertain as it depends on the unknown regulation trajectories ex-post of the market clearing. Addressing the nonconvexity and uncertainty in a multi-interval co-optimized real-time energy-reserve market, we introduce a simple restriction on the SoC-dependent bids along with a robust optimization formulation, transforming the non-convex market clearing under uncertainty into a standard convex piece-wise linear program and making it possible for large-scale storage integration. Under reasonable assumptions, we show that SoC-dependent bids yield higher profit for storage participants than that from SoC-independent bids. Numerical simulations demonstrate a 28%-150% profit increase of the proposed SoC-dependent bids compared with the SoC-independent counterpart.

eess.SY↗

arXiv2024

SAM2-UNet: Segment Anything 2 Makes Strong Encoder for Natural and Medical Image Segmentation

Image segmentation plays an important role in vision understanding. Recently, the emerging vision foundation models continuously achieved superior performance on various tasks. Following such success, in this paper, we prove that the Segment Anything Model 2 (SAM2) can be a strong encoder for U-shaped segmentation models. We propose a simple but effective framework, termed SAM2-UNet, for versatile image segmentation. Specifically, SAM2-UNet adopts the Hiera backbone of SAM2 as the encoder, while the decoder uses the classic U-shaped design. Additionally, adapters are inserted into the encoder to allow parameter-efficient fine-tuning. Preliminary experiments on various downstream tasks, such as camouflaged object detection, salient object detection, marine animal segmentation, mirror detection, and polyp segmentation, demonstrate that our SAM2-UNet can simply beat existing specialized state-of-the-art methods without bells and whistles. Project page: \url{https://github.com/WZH0120/SAM2-UNet}.

cs.CV↗