arXiv subjects

Zhiyuan Li

Publications and source records attributed to Zhiyuan Li.

At least 55 records · Page 3Linked to original sources

arXiv2026

Constraints on the Hot Circumgalactic Medium around Nearby L* Galaxies from SRG/eROSITA All Sky Survey

The circumgalactic medium (CGM), a multi-phase gas acting as the dynamic interface between the main body of a normal galaxy and the intergalactic medium, is a key ingredient of the galactic ecosystem and provides crucial diagnostics to a wide array of physical processes affecting galaxy evolution. However, direct evidence for a predominantly hot (at a characteristic temperature of a few million Kelvin) CGM around present-day $L^*$ galaxies remains elusive, despite extensive observational searches and strong physical reasons that it should exist. Here, we present a systematic search for the hot CGM around a representative sample of nearby $L^*$ galaxies selected from the 50 Mpc Galaxy Catalog (50MGC), by stacking their X-ray images and spectra from the SRG/eROSITA all-sky survey. Significant ($7.4σ$) diffuse X-ray emission is detected out to a galactocentric radius $\gtrsim$ 50 kpc, the cumulative spectrum of which requires a substantial thermal component, consistent with a hot plasma with log-normal temperature distribution, and thus argues against a predominantly non-thermal origin. The diffuse emission exhibits a relatively steep intensity profile ($β=0.52_{-0.06}^{+0.08}$), inferring a moderate amount ($2\times10^{10}~M_{\odot}$) of hot gas within 10--200 kpc. The radial distribution and total amount of the hot CGM gas are broadly in agreement with prediction by IllustrisTNG simulations for present-day Milky Way-like galaxies, while also showing dependencies on both star formation and nuclear activity. The constraints on the hot CGM derived in this study hold promise for calibrating key physical processes, such as stellar feedback and active galactic nucleus feedback, in next-generation cosmological simulations.

astro-ph.GA↗

arXiv2026

StageGuard: Learning Stage Transitions for Long-Horizon Robot Tasks via Agentic Distillation

Hierarchical planning frameworks combine skills from multiple robot control policies for long-horizon task execution, where determining when to terminate the current skill and advance to the next subtask is essential. Existing approaches often rely on pre-designed completion signal checkers that are hard to obtain in real-world execution. Large-scale vision-language models (VLMs) offer strong reasoning capabilities, but their decision boundaries are not inherently aligned with task completion criteria, while cloud deployment and lengthy reasoning introduce substantial latency, limiting real-time monitoring. We propose StageGuard, an agentic distillation framework for accurate and efficient stage-transition decisions. StageGuard combines teacher-model reasoning with demonstration trajectories to generate structured explanations of subtask completion and policy switching. A lightweight student VLM uses these explanations to generate compact self-explanations, which are used for supervised fine-tuning. We evaluate stage-transition prediction on trajectories from two benchmarks and assess closed-loop task success through integration into hierarchical robot control on BEHAVIOR-1K, with further validation on real robots. Results show substantial improvements in stage-transition prediction while supporting efficient online monitoring.

cs.RO↗

arXiv2026

Mem2Ego: Empowering Vision-Language Models with Global-to-Ego Memory for Long-Horizon Embodied Navigation

Recent advancements in Large Language Models (LLMs) and Vision-Language Models (VLMs) have made them powerful tools in embodied navigation, enabling agents to leverage commonsense and spatial reasoning for efficient exploration in unfamiliar environments. Existing LLM-based approaches convert global memory, such as semantic or topological maps, into language descriptions to guide navigation. While this improves efficiency and reduces redundant exploration, the loss of geometric information in language-based representations hinders spatial reasoning, especially in intricate environments. To address this, VLM-based approaches directly process ego-centric visual inputs to select optimal directions for exploration. However, relying solely on a first-person perspective makes navigation a partially observed decision-making problem, leading to suboptimal decisions in complex environments. In this paper, we present a novel vision-language model (VLM)-based navigation framework that addresses these challenges by adaptively retrieving task-relevant cues from a global memory module and integrating them with the agent's egocentric observations. By dynamically aligning global contextual information with local perception, our approach enhances spatial reasoning and decision-making in long-horizon tasks. The proposed method surpasses previous state-of-the-art approaches by a significant margin on both the HSSD and HM3D benchmarks and demonstrates strong performance on a real robot.

cs.RO↗

arXiv2026

A Scenario-Knowledge-Driven Pipeline for Just-in-Time Assistance

Detecting a silently struggling kiosk user is only the first step; deciding whether, when, and how to help depends on scenario knowledge usually buried in model weights and thresholds. We propose a scenario-knowledge-driven pipeline: a single scenario knowledge document, human-authored and version-controlled, configures sensing, constrains LLM reasoning, and shapes a graded intervention proposal. Narration, assistance-need assessment, and proposal are kept separate for independent audit. As proof of concept, we replay two recorded kiosk sessions offline, chosen before the runs for their struggle evidence and retrospective detail. Both cases support what the design promises: checkable reporting and measured escalation. Across 95 updates, every sentence of the append-only narration cites the primitive events underlying it, and the rule layer detects 12 of 13 and 7 of 7 annotated struggle episodes under a strict criterion. The assessor de-escalates on recovery and reaches the top rung exactly once, under maximally converging evidence. At the decisive help-seeking turn, narration, assessment, and the participants' retrospective accounts converge. The appropriateness of these interventions, the pipeline's restraint on sessions without struggle, and the document's transfer to a new scenario frame the agenda.

cs.HC↗

arXiv2026

Logarithmic stability for recovering initial data of fractional heat equations from thin-set observations

We study the inverse problem of recovering initial data for fractional heat equations on bounded domains from observations taken on two types of possibly Lebesgue-null sets: the first consists of general thin sets whose Hausdorff dimension exceeds $n-1$; the second consists of specially constructed sets of zero Hausdorff dimension, built from algebraic irrational points and rapidly accumulating sequences.By extending thin-set observability inequalities from the classical heat equation to the fractional setting and employing a unified spectral inequality with exponent $β\in(0,s)$, we establish a quantitative observability estimate for every fractional exponent $s>1/2$. For general thin sets one has $β=1/2$; for the special zero-dimensional sets any $β\in(1/2,s)$ is admissible, yet the short-time observability cost always remains of exponential type, with the rate governed by $β$ relative to $s$. Under an a priori smoothness assumption, we also prove a logarithmic stability estimate.We further design a regularised least-squares reconstruction algorithm and provide a conditional convergence analysis, showing that the reconstruction error is bounded by the sum of a spectral truncation term and a logarithmic term dictated by the continuous stability, and that the error decays logarithmically as the noise level tends to zero. Numerical simulations using very few observation points confirm the feasibility of the proposed approach.

math.AP↗

arXiv2026

Octopus v2: On-device language model for super agent

Language models have shown effectiveness in a variety of software applications, particularly in tasks related to automatic workflow. These models possess the crucial ability to call functions, which is essential in creating AI agents. Despite the high performance of large-scale language models in cloud environments, they are often associated with concerns over privacy and cost. Current on-device models for function calling face issues with latency and accuracy. Our research presents a new method that empowers an on-device model with 2 billion parameters to surpass the performance of GPT-4 in both accuracy and latency, and decrease the context length by 95\%. When compared to Llama-7B with a RAG-based function calling mechanism, our method enhances latency by 35-fold. This method reduces the latency to levels deemed suitable for deployment across a variety of edge devices in production environments, aligning with the performance requisites for real-world applications.

cs.CL↗

arXiv2026

Verify to Amplify: Improving Reasoning via Learned Chain-of-Thought Verification

Large Language Models (LLMs) using chain-of-thought have demonstrated great potential for solving complex reasoning and planning tasks. Despite these advances, LLM-generated outputs remain susceptible to errors, making verification important for reliable reasoning systems. Learned verifiers can increase trust, enforce safety constraints, and ensure alignment with personal preferences, while also providing feedback to improve generation. This raises a central challenge: when learned verifiers are used to guide generation, the feedback loop between generator and verifier may induce a distribution shift. This is particularly salient for process reward models, a prominent class of learned verifiers that score or classify individual steps in a chain-of-thought reasoning trace. Motivated by this challenge, we propose a new online learning framework for chain-of-thought verifiers that, given a problem statement and a reasoning trace, check the correctness of each reasoning step given the preceding steps. Highlighting the asymmetric role of soundness errors (accepting an incorrect reasoning step) and completeness errors (flagging a correct step as wrong), we introduce novel notions of dimension that characterize their optimal tradeoff. We then show how our learned verifiers can boost the accuracy of a weak generator. Assuming that the generator can produce a correct next step with a small success probability, we show how to learn a strong generator with small error and abstention rates. Our results also allow learning from offline data when queries to an expert verifier can be simulated from a small set of correct reasoning traces. However, we establish a separation between our approach and learning from offline expert demonstrations: we show that learning from offline demonstrations cannot in general achieve the soundness-completeness guarantees produced by our interactive learning approach.

cs.LG↗

arXiv2026

Search for neutrinoless quadruple beta decay of $^{136}$Xe in PandaX-4T detector

The observation of neutrinoless quadruple beta decay (0$ν$4$β$) in the absence of neutrinoless double beta decay (0$ν$2$β$) has been argued to provide a strong indication that neutrinos are Dirac particles. We report a search for 0$ν$4$β$ decay of $^{136}\text{Xe}$ using a total $^{136}\text{Xe}$ exposure of 148.4 kg$\cdot$yr, collected during the commissioning and the first science runs of the PandaX-4T experiment. No significant excess of events over the background is observed. A lower limit on the 0$ν$4$β$ decay half-life of $^{136}\text{Xe}$ is set at 6.01 x $10^{24}$ yr at the 90% confidence level. This result establishes the most stringent constraint on this process in xenon, demonstrating the unique capability of the PandaX-4T detector in probing lepton number violation and shedding light on the fundamental nature of neutrinos.

nucl-ex↗

arXiv2026

Inverse Heat Source Problems from Boundary Flux and Interior Observations on Sets of Low Hausdorff Dimension

This paper investigates conditional stability for inverse source problems for the heat equation with a known temporal factor and an unknown spatial component in a bounded $C^{1,1}$ domain. We focus on observations supported on sets of low Hausdorff dimension and establish conditional stability in this setting. For boundary observations on compact sets of positive $q$-dimensional Hausdorff content, we establish logarithmic stability from full-time boundary flux observations and double-logarithmic stability from delayed-time boundary flux observations. The admissible dimensional ranges are $q>d-2$ when the observation set is contained in a flat boundary patch and $q>d-1-c_{d+1}$ on a general $C^{1,1}$ boundary, where $c_{d+1}>0$ depends only on the dimension. A key ingredient in deriving these results is a new boundary spectral inequality for the Dirichlet Laplacian, which controls a finite Dirichlet spectral sum through observations of the normal derivative of its elliptic extension on such a boundary set. Our results also cover inverse heat source problems with interior observations on sets of positive $q$-dimensional Hausdorff content for some $q>d-1$, yielding logarithmic stability from full-time observations for general sources in $H_0^1(Ω)$ and Hölder stability from terminal-time observations for sources in a suitable spectral Gevrey class.

math.SP↗

arXiv2026

Origin of the tetragonal-to-hexagonal phase transitions in Fe-doped BaTiO$_3$

Based on detailed first-principles calculations, we investigate the tetragonal-to-hexagonal phase transition in Fe-doped BaTiO$_3$. Free energy calculations confirm a crossover from the tetragonal to hexagonal phases at 2.7--6\% Fe on cooling from the sintering temperature, in agreement with experimental observations, where comparative calculations show that neither CaTiO$_3$ nor SrTiO$_3$ exhibits similar behavior under equivalent substitution. Furthermore, three possible mechanisms are quantified: oxygen vacancies shift the crossover concentration from 2.7--6\% to 1.1--4.7\%, Jahn-Teller distortions are confined to the tetragonal phase, both favoring tetragonal-to-hexagonal phase transitions; whereas the tolerance factor is reduced in comparison with that of pristine BaTiO$_3$ for reasonable Fe valence states, disfavoring the occurrence of the hexagonal phases. Detailed analysis on the electronic structure reveals that the charge redistribution induced by oxygen vacancy is strongly orbital dependent, the added charge entering the $d_{z^2}$ orbital.

cond-mat.mtrl-sci↗

arXiv2026

EditaLive! Unified Character Video Editing for Live Streaming

Conventional video editing primarily focuses on scene-level content, whereas live streaming places greater emphasis on the human subject. However, directly applying existing video-editing methods to human-centric live streaming remains challenging, as they may introduce facial-expression inconsistencies and typically depend on multiple offline inference steps, making them unsuitable for real-time interaction. We propose EditaLive, a novel framework for real-time streaming character video editing. In detail, we start from a pretrained image animation model (Wan-Animate), which naturally decouples appearance from motion, and repurpose it as the base model for instruction-based human-centric video editing by reference frame editing and video reconstruction via the collected CharEdit-50K dataset. Besides, we adapt the model from offline bidirectional to causal streaming generation, and design an aligned self-rollout distillation strategy that compresses the model into a two-step sampler, where fixed RoPE and align forcing reduce training--inference discrepancies, and first-frame preserved sparse attention filters redundant historical information to mitigate appearance drift. Extensive experiments demonstrate that EditaLive delivers state-of-the-art editing performance with faithful preservation of facial expressions and low-latency real-time streaming inference.

cs.CV↗

arXiv2026

CaSKG: Counterfactual-Causal Skill Graphs for Scalable Agent Skill Retrieval

Reusable skill libraries allow large language model (LLM) agents to reuse procedural knowledge across tasks, but they also turn memory access into a challenging retrieval problem. Full-library prompting preserves coverage at high context cost, vector retrieval returns compact neighborhoods but treats skills as independent text, and graph-based retrieval can recover workflow context only when the edges that carry relevance are reliable. We propose CaSKG, a counterfactual-causal skill graph framework that calibrates procedural relations before retrieval. CaSKG first builds a high-recall directed candidate graph from semantic, lexical, input/output, and structural evidence, with repair evidence and an optional LLM judge further refining candidate scores. It then applies direction-conditioned textual counterfactual probes that remove, substitute, and reorder skill pairs, aggregates the evidence with Bayesian smoothing, and publishes a state-filtered weighted graph for task-conditioned expansion. The graph is constructed offline and used without changing the downstream agent policy or task interface. Across six LLM backbones on ALFWorld ID-140 and ScienceWorld U211, CaSKG achieves the highest task score in all twelve combinations of model and benchmark. Relative to Graph-of-Skills (GoS), it improves the six-model macro-average ScienceWorld score from 72.62 to 80.50 and ALFWorld success from 80.01\% to 86.79\%, while reducing mean environment steps on both benchmarks. Qualitative and ablation analyses further show that calibrated edges help retrieval preserve prerequisites, state-changing actions, verification routines, and final completion steps. These results position edge-confidence calibration as an effective route to compact and executable skill retrieval at scale\footnote{Code is available at: https://github.com/ZhiyuanLi218/Caskg }.

cs.AI↗

arXiv2026

LG-GER: Language-Guided Group Emotion Recognition via Multimodal Evidence Distillation

Inferring the collective emotional state of a group of people from a single image, a task known as group emotion recognition (GER), requires integrating spatially distributed cues such as faces, poses, interactions, and scene context. Current methods rely on detector-driven multi-stream pipelines. These are trained with only image-level supervision that lacks guidance on which regions matter or how strongly each contributes. We propose LG-GER, a language-guided distillation framework that uses a multimodal large language model (MLLM) to generate dense, spatially grounded evidence, i.e., bounding boxes paired with emotion signals and confidence scores, for the training images. This structured evidence is distilled into a single vision-language model (VLM) backbone through four complementary losses: classification, region-text grounding, spatial emotion, and spatial confidence regression. At inference, LG-GER requires no detectors, no MLLM, and no multi-stream fusion, making GER practical for real-time and resource-constrained deployment. LG-GER has been evaluated on two benchmark GER datasets (GroupEmoW and GAF~3.0) and achieves competitive or superior results compared to state-of-the-art methods that require detection and multi-stream processing at inference.

cs.CV↗

arXiv2026

ARC-Bench: Closed-Loop Replanning Masks Broken Action Ranking in Frozen JEPA World Models

Reward-free latent world models plan by scoring candidate actions with distances in a frozen latent space: an action is preferred if its predicted future embedding lands closer to the goal embedding. This silently assumes that latent closeness is action-rankable, i.e., that ordering candidates by latent distance agrees with ordering them by true cost. We audit this assumption directly. We introduce ARC-Bench, a no-leak, fixed-candidate protocol that measures whether frozen JEPA-style objectives rank candidate actions correctly, and apply it to official released JEPA-WM checkpoints across navigation and manipulation-style control. The assumption fails, severely and structurally: on the official manipulation audits the top-scored candidate is almost always suboptimal, and the same inversion appears in the maze domains. A controlled visual-backbone extension shows that the defect persists when DINOv2 is replaced by video-pretrained V-JEPA 1 and V-JEPA 2 encoders at ViT-L/ViT-G scale. Provenance, undertraining, matched-budget backbone controls, and metric-circularity controls rule out trivial explanations. We then explain why this defect has stayed invisible: closed-loop replanning masks it. When we reduce the planner's replanning frequency, success collapses in both a navigation and a manipulation domain, and the episodes rescued by frequent replanning are enriched for severe first-plan ranking failures in the PointMaze first-plan diagnostic. Closed-loop success rates therefore systematically overstate the rankability of frozen latent representations. ARC-Bench supplies the measurement, and the masking mechanism the explanation, for methods that adapt, amortize, or replan around latent-space planners without directly auditing released JEPA-WM action rankability.

cs.AI↗

arXiv2026

Spiking Local Interaction and Adaptive Complementary Fusion for Spiking Transformer

Spiking Transformers model token interactions primarily through spiking self-attention (SSA). However, binary query and key representations map continuous similarities to sparse and discrete relation responses, which may suppress weak relations and limit the propagation of local spatial context. To address this limitation, we introduce Spiking Local Interaction (SLI) and Adaptive Complementary Fusion (ACF). SLI establishes an attention-independent pathway for direct information exchange among neighboring spiking tokens using lightweight depthwise--pointwise transformations. ACF integrates SSA and SLI through layer-specific, channel-wise coefficients that adaptively balance their contributions at different network depths. The proposed design preserves the original attention formulation and can be incorporated into different Spiking Transformer architectures with modest parameter overhead. Experiments on ImageNet-1K, CIFAR-10, CIFAR-100, CIFAR10-DVS, and ADE20K show consistent improvements across image classification, event-based recognition, and semantic segmentation. In particular, QKFormer with SLI and ACF achieves $84.37\%$ Top-1 accuracy on ImageNet-1K and $37.5\%$ mIoU on ADE20K, where the segmentation model is trained without ImageNet pretraining. Ablation studies and qualitative analyses further indicate that SSA and SLI capture complementary interaction patterns and that learnable fusion consistently outperforms fixed weighting.

cs.NE↗

arXiv2026

NTIRE 2026 Challenge on Single Image Reflection Removal in the Wild: Datasets, Results, and Methods

In this paper, we review the NTIRE 2026 challenge on single-image reflection removal (SIRR) in the wild. SIRR is a fundamental task in image restoration. Despite progress in academic research, most methods are tested on synthetic images or limited real-world images, creating a gap in real-world applications. In this challenge, we provide participants with the OpenRR-5k dataset. This dataset requires participants to process real-world images covering a range of reflection scenarios and intensities, aiming to generate clean images without reflections. The challenge attracted more than 100 registrations, with eleven of them participating in the final testing phase. The top-ranked methods advanced the state-of-the-art reflection removal performance and earned unanimous recognition from five experts in the field. The proposed OpenRR-5k dataset is available at https://huggingface.co/datasets/qiuzhangTiTi/OpenRR-5k, and the homepage of this challenge is at https://github.com/caijie0620/OpenRR-5k.

cs.CV↗

arXiv2026

LiveLight: Real-time Streaming Video Relighting with Interactive Control

We present LiveLight, the first diffusion-based framework for real-time streaming video relighting with interactive 3D lighting control. Achieving this is non-trivial, as it requires overcoming three critical challenges: effectively injecting dynamic 3D lighting into a diffusion model, maintaining high-fidelity generation under an extremely low NFE (Number of Function Evaluations) budget for real-time speed, and facilitating continuous streaming for interactive control. To address these pain points, we propose three key designs. First, for accurate lighting injection, we propose a lightweight adapter that feeds Multi-Plane Light Irradiance (MPLI) conditions-depth-aware irradiance maps encoding 3D lighting geometry-directly into the diffusion backbone. Second, to prevent rendering quality degradation at low NFEs towards real-time distillation, we introduce a geometry-guided feedback branch. This training-time constraint leverages a frozen geometry estimator to enforce depth- and normal-consistent relighting, ensuring geometrically plausible shading without adding inference overhead. Finally, to enable streaming interaction, we develop a progressive rolling-window strategy that maintains a denoising ladder of latent chunks at varying noise levels. By propagating intermediate states, this strategy guarantees temporal coherence and supports arbitrarily long video relighting with per-frame reference refresh. Extensive experiments on real-world and synthetic benchmarks demonstrate that LiveLight achieves state-of-the-art relighting quality while running at real-time speed, significantly outperforming offline baselines in temporal stability, lighting controllability, and user preference. To foster real-time interactive relighting research, we will publicly release our models, training data, and synthetic data generator.

cs.CV↗

arXiv2026

LADY: Linear Attention for Autonomous Driving Efficiency without Transformers

End-to-end autonomous driving has emerged as a promising paradigm. However, state-of-the-art methods rely heavily on Transformer architectures. The inherent quadratic complexity of Transformers restricts their ability to model long-range spatial and temporal dependencies, particularly on resource-constrained edge platforms. Given the inherent demand for efficient temporal modeling in autonomous driving, this computational bottleneck severely constrains real-time deployment. While linear attention mechanisms offer a computationally efficient alternative, existing architectures are predominantly limited to self-attention, lacking the cross-modal capabilities essential for autonomous driving. In this work, we propose LADY, the first fully linear attention-based generative model for end-to-end autonomous driving. LADY incorporates a novel, lightweight linear cross-attention (LICA) mechanism to enable effective cross-modal interaction while preserving linearity. A key advantage of our framework is its ability to fuse long-range temporal contexts during inference with constant computational and memory costs ($O(1)$), regardless of the historical sequence length. Experiments on the NAVSIM and Bench2Drive benchmarks demonstrate that LADY achieves performance comparable to state-of-the-art methods, delivering competitive planning accuracy with significantly reduced latency. Furthermore, efficiency benchmarking on edge devices validates the model's feasibility for resource-limited scenarios.

cs.AI↗