arXiv subjects

Rui Li

Publications and source records attributed to Rui Li.

At least 145 records · Page 8Linked to original sources

arXiv2026

Pressure-induced superconductivity beyond magnetic quantum criticality in a Kondo ferromagnet

Quantum phase transitions are an established setting for emergent phenomena driven by strong electronic correlations, including strange metals and unconventional superconductivity. These have been explored extensively in Kondo lattice materials tuned to an antiferromagnetic quantum critical point (QCP), but superconductivity emerging near ferromagnetic quantum criticality is not yet observed, and the conditions under which it occurs in proximity to ferromagnetism are undetermined. Here, we report a new setting for superconductivity in the ferromagnetic Kondo-lattice material Ce5CoGe2, where there is a ferromagnetic ground state at ambient pressure, which evolves to antiferromagnetism under applied pressures. The antiferromagnetic transition is suppressed to a zero-temperature QCP, which is accompanied by strange-metal behavior. Superconductivity does not occur at the QCP, but instead appears at pressures beyond the magnetic instability. These findings suggest that Ce5CoGe2 represents a distinct class of correlated materials exhibiting a unique scenario for the emergence of superconductivity, likely associated with unconventional pairing mechanisms beyond spin-fluctuations.

cond-mat.supr-con↗

arXiv2026

UniVTAC: A Unified Simulation Platform for Visuo-Tactile Manipulation Data Generation, Learning, and Benchmarking

Robotic manipulation has seen rapid progress with vision-language-action (VLA) policies. However, visuo-tactile perception is critical for contact-rich manipulation, as tasks such as insertion are difficult to complete robustly using vision alone. At the same time, acquiring large-scale and reliable tactile data in the physical world remains costly and challenging, and the lack of a unified evaluation platform further limits policy learning and systematic analysis. To address these challenges, we propose UniVTAC, a simulation-based visuo-tactile data synthesis platform that supports three commonly used visuo-tactile sensors and enables scalable and controllable generation of informative contact interactions. Based on this platform, we introduce the UniVTAC Encoder, a visuo-tactile encoder trained on large-scale simulation-synthesized data with designed supervisory signals, providing tactile-centric visuo-tactile representations for downstream manipulation tasks. In addition, we present the UniVTAC Benchmark, which consists of eight representative visuo-tactile manipulation tasks for evaluating tactile-driven policies. Experimental results show that integrating the UniVTAC Encoder improves average success rates by 17.1% on the UniVTAC Benchmark, while real-world robotic experiments further demonstrate a 25% improvement in task success. Our webpage is available at https://univtac.github.io/.

cs.RO↗

arXiv2026

Galaxy Light profile neural Networks (GaLNets). II. Bulge-Disc decomposition in optical space-based observations

Bulge-disk (B-D) decomposition is an effective diagnostic to characterize the galaxy morphology and understand its evolution across time. So far, high-quality data have allowed detailed B-D decomposition to redshift below 0.5, with limited excursions over small volumes at higher redshifts. Next-generation large sky space surveys in optical, e.g. from the China Space Station Telescope (CSST), and near-infrared, e.g. from the space EUCLID mission, will produce a gigantic leap in these studies as they will provide deep, high-quality photometric images over more than 15000 deg2 of the sky, including billions of galaxies. Here, we extend the use of the Galaxy Light profile neural Network (GaLNet) to predict 2-Sérsic model parameters, specifically from CSST data. We simulate point-spread function (PSF) convolved galaxies, with realistic B-D parameter distributions, on CSST mock observations to train the new GaLNet and predict the structural parameters (e.g. magnitude, effective radius, Sersic index, axis ratio, etc.) of both bulge and disk components. We find that the GaLNet can achieve very good accuracy for most of the B-D parameters down to an $r$-band magnitude of 23.5 and redshift $\sim$1. The best accuracy is obtained for magnitudes, implying accurate bulge-to-total (B/T) estimates. To further forecast the CSST performances, we also discuss the results of the 1-Sérsic GaLNet and show that CSST half-depth data will allow us to derive accurate 1-component models up to $r\sim$24 and redshift z$\sim$1.7.

astro-ph.GA↗

arXiv2026

Towards Adaptive, Scalable, and Robust Coordination of LLM Agents: A Dynamic Ad-Hoc Networking Perspective

Multi-agent architectures built on large language models (LLMs) have demonstrated the potential to realize swarm intelligence through well-crafted collaboration. However, the substantial burden of manual orchestration inherently raises an imperative to automate the design of agentic workflows. We frame such an agent coordination challenge as a classic problem in dynamic ad-hoc networking: How to establish adaptive and reliable communication among a scalable number of agentic hosts? In response to this unresolved dilemma, we introduce RAPS, a reputation-aware publish-subscribe paradigm for adaptive, scalable, and robust coordination of LLM agents. RAPS is grounded in the Distributed Publish-Subscribe Protocol, allowing LLM agents to exchange messages based on their declared intents rather than predefined topologies. Beyond this substrate, RAPS further incorporates two coherent overlays: (i) Reactive Subscription, enabling agents to dynamically refine their intents; and (ii) Bayesian Reputation, empowering each agent with a local watchdog to detect and isolate malicious peers. Extensive experiments over five benchmarks showcase that our design effectively reconciles adaptivity, scalability, and robustness in a unified multi-agent coordination framework.

cs.AI↗

arXiv2026

TeleBoost: A Systematic Alignment Framework for High-Fidelity, Controllable, and Robust Video Generation

Post-training is the decisive step for converting a pretrained video generator into a production-oriented model that is instruction-following, controllable, and robust over long temporal horizons. This report presents a systematical post-training framework that organizes supervised policy shaping, reward-driven reinforcement learning, and preference-based refinement into a single stability-constrained optimization stack. The framework is designed around practical video-generation constraints, including high rollout cost, temporally compounding failure modes, and feedback that is heterogeneous, uncertain, and often weakly discriminative. By treating optimization as a staged, diagnostic-driven process rather than a collection of isolated tricks, the report summarizes a cohesive recipe for improving perceptual fidelity, temporal coherence, and prompt adherence while preserving the controllability established at initialization. The resulting framework provides a clear blueprint for building scalable post-training pipelines that remain stable, extensible, and effective in real-world deployment settings.

cs.CV↗

arXiv2026

State-Augmented Graphs for Circular Economy Triage

Circular economy (CE) triage is the assessment of products to determine which sustainable pathway they can follow once they reach the end of their usefulness as they are currently being used. Effective CE triage requires adaptive decisions that balance retained value against the costs and constraints of processing and labour. This paper presents a novel decision-making framework as a simple deterministic solver over a state-augmented Disassembly Sequencing Planning (DSP) graph. By encoding the disassembly history into the state, our framework enforces the Markov property, enabling optimal, recursive evaluation by ensuring each decision only depends on the previous state. The triage decision involves choices between continuing disassembly or committing to a CE option. The model integrates condition-aware utility based on diagnostic health scores and complex operational constraints. We demonstrate the framework's flexibility with a worked example: the hierarchical triage of electric vehicle (EV) batteries, where decisions are driven by the recursive valuation of components. The example illustrates how a unified formalism enables the accommodation of varying mechanical complexity, safety requirements, and economic drivers. This unified formalism therefore provides a tractable and generalisable foundation for optimising CE triage decisions across diverse products and operational contexts.

cs.AI↗

arXiv2026

TeachBench: A Syllabus-Grounded Framework for Evaluating Teaching Ability in Large Language Models

Large language models (LLMs) show promise as teaching assistants, yet their teaching capability remains insufficiently evaluated. Existing benchmarks mainly focus on problem-solving or problem-level guidance, leaving knowledge-centered teaching underexplored. We propose a syllabus-grounded evaluation framework that measures LLM teaching capability via student performance improvement after multi-turn instruction. By restricting teacher agents to structured knowledge points and example problems, the framework avoids information leakage and enables reuse of existing benchmarks. We instantiate the framework on Gaokao data across multiple subjects. Experiments reveal substantial variation in teaching effectiveness across models and domains: some models perform well in mathematics, while teaching remains challenging in physics and chemistry. We also find that incorporating example problems does not necessarily improve teaching, as models often shift toward example-specific error correction. Overall, our results highlight teaching ability as a distinct and measurable dimension of LLM behavior.

cs.AI↗

arXiv2026

Self-Compression of Chain-of-Thought via Multi-Agent Reinforcement Learning

The inference overhead induced by redundant reasoning undermines the interactive experience and severely bottlenecks the deployment of Large Reasoning Models. Existing reinforcement learning (RL)-based solutions tackle this problem by coupling a length penalty with outcome-based rewards. This simplistic reward weighting struggles to reconcile brevity with accuracy, as enforcing brevity may compromise critical reasoning logic. In this work, we address this limitation by proposing a multi-agent RL framework that selectively penalizes redundant chunks, while preserving essential reasoning logic. Our framework, Self-Compression via MARL (SCMA), instantiates redundancy detection and evaluation through two specialized agents: \textbf{a Segmentation Agent} for decomposing the reasoning process into logical chunks, and \textbf{a Scoring Agent} for quantifying the significance of each chunk. The Segmentation and Scoring agents collaboratively define an importance-weighted length penalty during training, incentivizing \textbf{a Reasoning Agent} to prioritize essential logic without introducing inference overhead during deployment. Empirical evaluations across model scales demonstrate that SCMA reduces response length by 11.1\% to 39.0\% while boosting accuracy by 4.33\% to 10.02\%. Furthermore, ablation studies and qualitative analysis validate that the synergistic optimization within the MARL framework fosters emergent behaviors, yielding more powerful LRMs compared to vanilla RL paradigms.

cs.AI↗

arXiv2026

The SAGES Critical View of Safety Challenge: A Global Benchmark for AI-Assisted Surgical Quality Assessment

Advances in artificial intelligence (AI) for surgical quality assessment promise to democratize access to expertise, with applications in training, guidance, and accreditation. This study presents the SAGES Critical View of Safety (CVS) Challenge, the first AI competition organized by a surgical society, using the CVS in laparoscopic cholecystectomy, a universally recommended yet inconsistently performed safety step, as an exemplar of surgical quality assessment. A global collaboration across 54 institutions in 24 countries engaged hundreds of clinicians and engineers to curate 1,000 videos annotated by 20 surgical experts according to a consensus-validated protocol. The challenge addressed key barriers to real-world deployment in surgery, including achieving high performance, capturing uncertainty in subjective assessment, and ensuring robustness to clinical variability. To enable this scale of effort, we developed EndoGlacier, a framework for managing large, heterogeneous surgical video and multi-annotator workflows. Thirteen international teams participated, achieving up to a 17% relative gain in assessment performance, over 80% reduction in calibration error, and a 17% relative improvement in robustness over the state-of-the-art. Analysis of results highlighted methodological trends linked to model performance, providing guidance for future research toward robust, clinically deployable AI for surgical quality assessment.

cs.CV↗

arXiv2026

BootSeer: Analyzing and Mitigating Initialization Bottlenecks in Large-Scale LLM Training

Large Language Models (LLMs) have become a cornerstone of modern AI, driving breakthroughs in natural language processing and expanding into multimodal jobs involving images, audio, and video. As with most computational software, it is important to distinguish between ordinary runtime performance and startup overhead. Prior research has focused on runtime performance: improving training efficiency and stability. This work focuses instead on the increasingly critical issue of startup overhead in training: the delay before training jobs begin execution. Startup overhead is particularly important in large, industrial-scale LLMs, where failures occur more frequently and multiple teams operate in iterative update-debug cycles. In one of our training clusters, more than 3.5% of GPU time is wasted due to startup overhead alone. In this work, we present the first in-depth characterization of LLM training startup overhead based on real production data. We analyze the components of startup cost, quantify its direct impact, and examine how it scales with job size. These insights motivate the design of Bootseer, a system-level optimization framework that addresses three primary startup bottlenecks: (a) container image loading, (b) runtime dependency installation, and (c) model checkpoint resumption. To mitigate these bottlenecks, Bootseer introduces three techniques: (a) hot block record-and-prefetch, (b) dependency snapshotting, and (c) striped HDFS-FUSE. Bootseer has been deployed in a production environment and evaluated on real LLM training workloads, demonstrating a 50% reduction in startup overhead.

cs.LG↗

arXiv2026

High-contrast double Bragg interferometry via detuning control

We propose high-contrast Mach-Zehnder atom interferometers based on double Bragg diffraction (DBD) operating under external acceleration. To mitigate differential Doppler shifts and experimental imperfections, we introduce a tri-frequency laser scheme with dynamic detuning control. We evaluate four detuning-control strategies-conventional DBD, constant detuning, linear detuning sweep (DS-DBD), and a hybrid protocol combining detuning sweep with optimal control theory (OCT)-using exact numerical simulations and a five-level S-matrix model. The OCT strategy provides the highest robustness, maintaining contrast above 95\% under realistic conditions, while the DS-DBD strategy sustains contrast above 90\% for well-collimated Bose-Einstein condensates. These results offer practical pathways to high-contrast, large-momentum-transfer DBD-based interferometers for precision quantum sensing and fundamental physics tests.

quant-ph↗

arXiv2026

Advances and Innovations in the Multi-Agent Robotic System (MARS) Challenge

Recent advancements in multimodal large language models and vision-languageaction models have significantly driven progress in Embodied AI. As the field transitions toward more complex task scenarios, multi-agent system frameworks are becoming essential for achieving scalable, efficient, and collaborative solutions. This shift is fueled by three primary factors: increasing agent capabilities, enhancing system efficiency through task delegation, and enabling advanced human-agent interactions. To address the challenges posed by multi-agent collaboration, we propose the Multi-Agent Robotic System (MARS) Challenge, held at the NeurIPS 2025 Workshop on SpaVLE. The competition focuses on two critical areas: planning and control, where participants explore multi-agent embodied planning using vision-language models (VLMs) to coordinate tasks and policy execution to perform robotic manipulation in dynamic environments. By evaluating solutions submitted by participants, the challenge provides valuable insights into the design and coordination of embodied multi-agent systems, contributing to the future development of advanced collaborative AI systems.

cs.RO↗

arXiv2026

FUSAR-KLIP: Towards Multimodal Foundation Models for Remote Sensing

Cross-modal artificial intelligence, represented by visual language models, has achieved significant success in general image understanding. However, a fundamental cognitive inconsistency exists between general visual representation and remote sensing image interpretation: remote sensing images couple topography, terrain, and spatial structure, thereby inherently requiring models to possess deep geoscientific understanding. This cognitive difference is further amplified in synthetic aperture radar (SAR) imagery: while SAR possesses irreplaceable all-weather, all-day observation capabilities, it is constrained by coherent imaging mechanisms, exhibiting significant modal heterogeneity with general images. To address this inconsistency, we propose FUSAR-KLIP, the first knowledge-guided general multimodal foundational model for SAR, along with reusable data and evaluation baselines. Specifically: (1) FUSAR-GEOVL-1M (the first large-scale SAR dataset with complete geographic projection attributes) was constructed, covering multiple satellite platforms, 120,000 images, and 135 cities; (2) Aligned structured text was generated through hierarchical cognitive thought chains, accurately encoding more than 1 million multidimensional semantic information from geomorphological environment and regional attributes to spatial relationships; (3) A self-consistent iterative optimization mechanism was designed to guide cross-modal learning with this knowledge information consistent with human cognition and physical laws in a self-supervised closed loop consisting of contrast, matching, and reconstruction; (4) A unified evaluation benchmark was established in 11 typical downstream tasks in the two major categories of vision and language, and compared with 15 mainstream foundation models.

cs.CV↗

arXiv2026

MIU2Net: weak-lensing mass inversion using deep learning with nested U-structures

One of the primary goals of next-generation gravitational lensing surveys is to measure the large-scale distribution of dark matter, which requires accurate mass inversion to convert weak-lensing shear maps into convergence (kappa) fields. This work develops a mass inversion method tailored for upcoming space missions such as CSST and Euclid, aiming to recover both the mass distribution and the convergence power spectrum with high fidelity. We introduce MIU2Net, a versatile deep-learning framework for kappa-map reconstruction based on the U2-Net architecture. A new loss function is constructed to jointly estimate the convergence field and its frequency-domain energy distribution, effectively balancing optimal mean squared error and optimal power-spectrum recovery. The method incorporates realistic observational effects into shear fields, including shape noise, reduced shear, and complex masks. Under noise levels anticipated for future space-based lensing surveys, MIU2Net recovers the convergence power spectrum with 4% uncertainties up to l approximately 500, significantly outperforming Wiener filtering and MCALens. Beyond two-point statistics, the method accurately reconstructs the convergence distribution, peak centroid, and peak amplitude. Compared to other learning-based approaches such as DeepMass, MIU2Net reduces the root-mean-square error by 5% without smoothing and by 38% with a 1-arcmin smoothing scale. MIU2Net represents a substantial advancement in mass inversion methodology, offering improved accuracy in both RMSE and power-spectrum reconstruction. It provides a promising tool for mapping dark matter environments and large-scale structures in the era of next-generation space lensing surveys.

astro-ph.CO↗

arXiv2026

Integrating Reinforcement Learning with Visual Generative Models: Foundations and Advances

Generative models have made significant progress in synthesizing visual content, including images, videos, and 3D/4D structures. However, they are typically trained with surrogate objectives such as likelihood or reconstruction loss, which often misalign with perceptual quality, semantic accuracy, or physical realism. Reinforcement learning (RL) offers a principled framework for optimizing non-differentiable, preference-driven, and temporally structured objectives. Recent advances demonstrate its effectiveness in enhancing controllability, consistency, and human alignment across generative tasks. This survey provides a systematic overview of RL-based methods for visual content generation. We review the evolution of RL from classical control to its role as a general-purpose optimization tool, and examine its integration into image, video, and 3D/4D generation. Across these domains, RL serves not only as a fine-tuning mechanism but also as a structural component for aligning generation with complex, high-level goals. We conclude with open challenges and future research directions at the intersection of RL and generative modeling.

cs.CV↗

arXiv2026

Be Your Own Red Teamer: Safety Alignment via Self-Play and Reflective Experience Replay

Large Language Models (LLMs) have achieved remarkable capabilities but remain vulnerable to adversarial ``jailbreak'' attacks designed to bypass safety guardrails. Current safety alignment methods depend heavily on static external red teaming, utilizing fixed defense prompts or pre-collected adversarial datasets. This leads to a rigid defense that overfits known patterns and fails to generalize to novel, sophisticated threats. To address this critical limitation, we propose empowering the model to be its own red teamer, capable of achieving autonomous and evolving adversarial attacks. Specifically, we introduce Safety Self- Play (SSP), a system that utilizes a single LLM to act concurrently as both the Attacker (generating jailbreaks) and the Defender (refusing harmful requests) within a unified Reinforcement Learning (RL) loop, dynamically evolving attack strategies to uncover vulnerabilities while simultaneously strengthening defense mechanisms. To ensure the Defender effectively addresses critical safety issues during the self-play, we introduce an advanced Reflective Experience Replay Mechanism, which uses an experience pool accumulated throughout the process. The mechanism employs a Upper Confidence Bound (UCB) sampling strategy to focus on failure cases with low rewards, helping the model learn from past hard mistakes while balancing exploration and exploitation. Extensive experiments demonstrate that our SSP approach autonomously evolves robust defense capabilities, significantly outperforming baselines trained on static adversarial datasets and establishing a new benchmark for proactive safety alignment.

cs.CR↗

arXiv2026

FMVP: Masked Flow Matching for Adversarial Video Purification

Video recognition models remain vulnerable to adversarial attacks, while existing diffusion-based purification methods suffer from inefficient sampling and curved trajectories. Directly regressing clean videos from adversarial inputs often fails to recover faithful content due to the subtle nature of perturbations; this necessitates physically shattering the adversarial structure. Therefore, we propose Flow Matching for Adversarial Video Purification FMVP. FMVP physically shatters global adversarial structures via a masking strategy and reconstructs clean video dynamics using Conditional Flow Matching (CFM) with an inpainting objective. To further decouple semantic content from adversarial noise, we design a Frequency-Gated Loss (FGL) that explicitly suppresses high-frequency adversarial residuals while preserving low-frequency fidelity. We design Attack-Aware and Generalist training paradigms to handle known and unknown threats, respectively. Extensive experiments on UCF-101 and HMDB-51 demonstrate that FMVP outperforms state-of-the-art methods (DiffPure, Defense Patterns (DP), Temporal Shuffling (TS) and FlowPure), achieving robust accuracy exceeding 87% against PGD and 89% against CW attacks. Furthermore, FMVP demonstrates superior robustness against adaptive attacks (DiffHammer) and functions as a zero-shot adversarial detector, attaining AUC-ROC scores of 0.98 for PGD and 0.79 for highly imperceptible CW attacks.

cs.CV↗

arXiv2026

HAPS: Hierarchical LLM Routing with Joint Architecture and Parameter Search

Large language model (LLM) routing aims to exploit the specialized strengths of different LLMs for diverse tasks. However, existing approaches typically focus on selecting LLM architectures while overlooking parameter settings, which are critical for task performance. In this paper, we introduce HAPS, a hierarchical LLM routing framework that jointly searches over model architectures and parameters. Specifically, we use a high-level router to select among candidate LLM architectures, and then search for the optimal parameters for the selected architectures based on a low-level router. We design a parameter generation network to share parameters between the two routers to mutually enhance their capabilities. In the training process, we design a reward-augmented objective to effectively optimize our framework. Experiments on two commonly used benchmarks show that HAPS consistently outperforms strong routing baselines. We have released our code at https://github.com/zihangtian/HAPS.

cs.CL↗