arXiv subjects

Yihao Hu

Publications and source records attributed to Yihao Hu.

At least 19 recordsLinked to original sources

arXiv2026

Towards Better Agents for Multi-Turn User Interaction: The Next User Turn Is More Than Context

User-facing tool agents must coordinate dialogue and tool use as user goals unfold over multiple turns. Yet interactive reinforcement learning typically reduces each rollout to a terminal reward, assigning the same credit to effective elicitation, errors, and later repair. The next user turn is more than context: it also provides noisy, temporally local evidence about the preceding user-to-user segment. We introduce \textbf{F}eedback-\textbf{A}ware \textbf{C}redit \textbf{A}ssignment (\textsc{FACA}), which aligns each reaction with that segment, derives a locally normalized reaction advantage, and adds it to verified terminal outcome advantage without an extra critic or rollout. Against an outcome-only Interactive GRPO control matched in simulator, visible dialogue, initialization, rollout, and optimization, \textsc{FACA} improves the nine-domain $\tau$-family average across three independently trained runs by 5.91 and 10.22 percentage points at 8B and 14B, respectively. Gains concentrate in Telecom; at 8B, randomizing reaction polarity removes the Telecom gain. The same ordering holds zero-shot on Pare-Bench and Co-Gym. These results demonstrate that next-turn user reactions provide actionable local credit for improving multi-turn user-interacting agents.

cs.AI↗

arXiv2026

Diagnosis Before Recovery: Turning Agent Failures into Selective Self-Correction

Self-correction is particularly useful when a failure constrains the next repair. Coding agents benefit from this property because compilers, tests, and execution traces turn many failures into typed recovery signals, but broad language-agent tasks often expose only a coarse task failure. This creates a tension for generic recovery playbooks: they broaden the agent's context precisely when the system needs a narrower repair interface, mixing incompatible signals for invalid actions, missing procedures, and strict-format errors. Our insight is that development-set failures can recover part of the missing diagnostic substrate by deciding which recovery interventions are admissible before test-time correction. We propose DARC, a diagnosis-guided recovery harness that profiles task-family failure modes, prunes mismatched interventions from a shared recovery library, and freezes a verifier-selected success-cost policy for deployment. This causal order makes correction selective: the harness first determines what kind of failure can be repaired, then decides how much recovery evidence to spend. In ALFWorld, AppWorld, and XBRL Finance, the same protocol yields an action-validity harness, a procedural-recovery fallback, and a format-precision retrieval policy; in each evaluated setting it improves average task performance over base agents and broad playbooks while reducing environment steps or retrieval budget. Our experiments show that failures need not trigger uniformly more context: DARC turns self-correction from prompt expansion into recovery-interface design. DARC provides a practical route toward more reliable agents in domains where compiler-like feedback is absent: making failures actionable before making contexts larger.

cs.CL↗

arXiv2026

I-Rex: An Interactive Debugger for SQL

SQL is declarative in nature and rich in its features. Writing semantically correct SQL queries and finding logical bugs in SQL are not easy, even for experienced programmers, who are often used to the mindset of working with general-purpose programming languages (GPLs). While there are many GPL debuggers, SQL debugging has received much less attention. In this paper, we present I-Rex, a SQL debugger that enables users to inspect the logical execution of SQL queries visually and interactively to identify and potentially fix logical bugs in the queries. I-Rex draws analogies to the debugging paradigm of GPLs (e.g., stepping, watchpoints, etc.), making it easier for programmers to adopt. However, unlike debugging GPLs, which involves executing the underlying program in full to the point of interest, I-Rex allows users to jump to arbitrary points of interest by leveraging the power of the database systems, through selective materialization and query rewrites. To simplify deployment, I-Rex acts as a lightweight middleware on top of the database system; it imposes no overhead to prepare a database for debugging and maintains no state in the database systems during debugging sessions. We demonstrate the effectiveness of I-Rex through performance experiments as well as a user study in an educational setting.

cs.DB↗

arXiv2026

CogSENet: Blind Image Deblurring with Blur-Conditioned Semantic Routing and Explicit Frequency Fusion

Blind image deblurring demands the recovery of high-fidelity details and coherent structures from complex, unknown degradations. Current blind image deblurring methods struggle with real-world, spatially varying degradations, and lack the semantic awareness necessary to reliably differentiate valid textures from artifacts. To bridge this gap, we propose CogSENet, a dynamic, semantic-aligned reconstruction framework inspired by the eagle's visual system. By mimicking the eagle's active saccadic scanning, we devise a Semantic-Driven State Space Module (SDSSM) with semantic-aware token regrouping via differentiable routing, enabling prompt-conditioned long-range dependency modeling. To ensure physically interpretable recovery of textures and structures, a BiFreqFusionBlock (BFFB) mirrors functional differentiation of the eagle's retina by decomposing features into high and low frequencies using wavelet transforms. Finally, we estimate a continuous Blur Field (CBF) from blur image and fuse it with CLIP semantic priors to modulate the deepest latent features, emulating focal adaptation and enabling adaptive restoration under spatially non-uniform blur. Extensive experiments demonstrate that CogSENetoutperforms state-of-the-art deblurring methods in both visual quality and structural fidelity with fewer parameters, while also performing favorably on dehazing, deraining, and denoising tasks.

cs.CV↗

arXiv2026

SEAL: Synergistic Co-Evolution of Agents and Learning Environments

Large Language Model (LLM) agents are increasingly improved through interaction, yet most self-evolution methods adapt either the policy or the learning environment in isolation. We identify this structural gap as \emph{Agent-Environment Misalignment}: the agent's capability frontier changes during training, while the environment that provides supervision remains static or only weakly coupled to the agent's revealed failures. We propose SEAL, a closed-loop co-evolution framework for interactive tool-use agents. SEAL collects on-policy trajectories under executable verification, diagnoses failed rollouts into turn-level failure labels, and uses these diagnoses as a shared signal for both environment-side adaptation and model-side policy optimization. The environment evolves its training-time learning interface by exposing clearer tool affordance cues, constraint information, and recovery-oriented feedback, while the policy is updated with diagnosis-guided advantage reweighting. Extensive experiments across in-distribution and out-of-distribution multi-turn tool-use evaluations show that SEAL improves low-resource agent learning: with only 400 training samples, it yields +8.25 to +26.25 average-point gains across three backbones and exhibits positive out-of-distribution transfer. These results demonstrate the value of jointly adapting the learner and its training-time learning substrate for robust self-improving LLM agents.

cs.CL↗

arXiv2026

AtlasVA: Self-Evolving Visual Skill Memory for Teacher-Free VLM Agents

Vision-language model (VLM) agents increasingly rely on memory-augmented reinforcement learning to reuse experience across long-horizon tasks, yet most existing frameworks store memory as text and depend on proprietary teacher models to summarize or refine it. This design is poorly matched to spatial decision making: geometric priors are compressed into lossy language, and sparse interaction is often supervised through delayed textual feedback rather than dense visually grounded signals. We argue that reusable experience for VLM agents should remain visually grounded. Based on this insight, we propose \textbf{AtlasVA}, a teacher-free visual skill memory framework that organizes memory into three complementary layers: spatial heatmaps, visual exemplars, and symbolic text skills. AtlasVA further evolves danger and affinity atlases directly from trajectory statistics and lightweight grid heuristics, and reuses these self-evolving atlases as potential-based shaping rewards for reinforcement learning. This unifies perception, memory, and optimization without external LLM supervision. Experiments on \textsc{Sokoban}, \textsc{FrozenLake}, 3D embodied navigation, and 3D robotic manipulation benchmarks show that AtlasVA consistently outperforms text-centric memory baselines and competitive VLM agents, with especially strong gains on spatially intensive tasks. Homepage: https://wangpan-ustc.github.io/AtlasvaWeb

cs.CV↗

arXiv2026

CFSR: Geometry-Conditioned Shadow Removal via Physical Disentanglement

Traditional shadow removal networks often treat image restoration as an unconstrained mapping, lacking the physical interpretability required to balance localized texture recovery with global illumination consistency. To address this, we propose CFSR, a multi-modal prior-driven framework that reframes shadow removal as a physics-constrained restoration process. By seamlessly integrating 3D geometric cues with large-scale foundation model semantics, CFSR effectively bridges the 2D-3D domain gap. Specifically, we first map observations into a custom HVI color space to suppress shadow-induced noise and robustly fuse RGB data with estimated depth priors. At its core, our Geometric & Semantic Dual Explicit Guided Attention mechanism utilizes DINO features and 3D surface normals to directly modulate the attention affinity matrix, structurally enforcing physical lighting constraints. To recover severely degraded regions, we inject holistic priors via a frozen CLIP encoder. Finally, our Frequency Collaborative Reconstruction Module (FCRM) achieves an optimal synthesis by decoupling the decoding process. Conditioned on geometric priors, FCRM seamlessly harmonizes the reconstruction of sharp high-frequency occlusion boundaries with the restoration of low-frequency global illumination. Extensive experiments demonstrate that CFSR achieves state-of-the-art performance across multiple challenging benchmarks.

cs.CV↗

arXiv2026

OmniVideo-R1: Reinforcing Audio-visual Reasoning with Query Intention and Modality Attention

While humans perceive the world through diverse modalities that operate synergistically to support a holistic understanding of their surroundings, existing omnivideo models still face substantial challenges on audio-visual understanding tasks. In this paper, we propose OmniVideo-R1, a novel reinforced framework that improves mixed-modality reasoning. OmniVideo-R1 empowers models to "think with omnimodal cues" by two key strategies: (1) query-intensive grounding based on self-supervised learning paradigms; and (2) modality-attentive fusion built upon contrastive learning paradigms. Extensive experiments on multiple benchmarks demonstrate that OmniVideo-R1 consistently outperforms strong baselines, highlighting its effectiveness and robust generalization capabilities.

cs.AI↗

arXiv2026

Dual Latent Memory for Visual Multi-agent System

While Visual Multi-Agent Systems (VMAS) promise to enhance comprehensive abilities through inter-agent collaboration, empirical evidence reveals a counter-intuitive "scaling wall": increasing agent turns often degrades performance while exponentially inflating token costs. We attribute this failure to the information bottleneck inherent in text-centric communication, where converting perceptual and thinking trajectories into discrete natural language inevitably induces semantic loss. To this end, we propose \textbf{L}$\mathbf{^{2}}$\textbf{-VMAS}, a novel model-agnostic framework that enables inter-agent collaboration with dual latent memories. Furthermore, we decouple the perception and thinking while dynamically synthesizing dual latent memories. Additionally, we introduce an entropy-driven proactive triggering that replaces passive information transmission with efficient, on-demand memory access. Extensive experiments among backbones, sizes, and multi-agent structures demonstrate that our method effectively breaks the "scaling wall" with superb scalability, improving average accuracy by 2.7-5.4% while reducing token usage by 21.3-44.8%.

cs.AI↗

arXiv2025

A Multi-Strategy Framework for Enhancing Shatian Pomelo Detection in Real-World Orchards

Shatian pomelo detection in orchards is essential for yield estimation and lean production, but models tuned to ideal datasets often degrade in practice due to device-dependent tone shifts, illumination changes, large scale variation, and frequent occlusion. We introduce STP-AgriData, a multi-scenario dataset combining real-orchard imagery with curated web images, and apply contrast/brightness augmentations to emulate unstable lighting. To better address scale and occlusion, we propose REAS-Det, featuring Global-Selective Visibility Convolution (GSV-Conv) that expands the visible feature space under global semantic guidance while retaining efficient spatial aggregation, plus C3RFEM, MultiSEAM, and Soft-NMS for refined separation and localization. On STP-AgriData, REAS-Det achieves 86.5% precision, 77.2% recall, 84.3% mAP@0.50, and 53.6% mAP@0.50:0.95, outperforming recent detectors and improving robustness in real orchard environments. The source code is available at: https://github.com/Genk641/REAS-Det.

cs.CV↗

arXiv2025

SDE-DET: A Precision Network for Shatian Pomelo Detection in Complex Orchard Environments

Pomelo detection is an essential process for their localization, automated robotic harvesting, and maturity analysis. However, detecting Shatian pomelo in complex orchard environments poses significant challenges, including multi-scale issues, obstructions from trunks and leaves, small object detection, etc. To address these issues, this study constructs a custom dataset STP-AgriData and proposes the SDE-DET model for Shatian pomelo detection. SDE-DET first utilizes the Star Block to effectively acquire high-dimensional information without increasing the computational overhead. Furthermore, the presented model adopts Deformable Attention in its backbone, to enhance its ability to detect pomelos under occluded conditions. Finally, multiple Efficient Multi-Scale Attention mechanisms are integrated into our model to reduce the computational overhead and extract deep visual representations, thereby improving the capacity for small object detection. In the experiment, we compared SDE-DET with the Yolo series and other mainstream detection models in Shatian pomelo detection. The presented SDE-DET model achieved scores of 0.883, 0.771, 0.838, 0.497, and 0.823 in Precision, Recall, mAP@0.5, mAP@0.5:0.95 and F1-score, respectively. SDE-DET has achieved state-of-the-art performance on the STP-AgriData dataset. Experiments indicate that the SDE-DET provides a reliable method for Shatian pomelo detection, laying the foundation for the further development of automatic harvest robots.

cs.CV↗

arXiv2025

Double Helix of atomic displacements in Ferroelectric PbTiO$_3$

Recent theoretical work has predicted the existence of a "dipole spiral" structure in strained freestanding membranes of PbTiO$_3$, suggesting a potential route to enhanced electromechanical responses [\href{https://journals.aps.org/prl/abstract/10.1103/PhysRevLett.133.046802}{PRL \textbf{133}, 046802 (2024)}]. However, its microscopic nature, energetic landscape, and electronic properties remain largely unexplored from a first-principles perspective. Here, using density function theory on PbTiO$_3$ under biaxial tensile strain, we identify a novel form of polar order: a chiral, non-collinear ferroelectric double helix. We find that the Pb- and Ti-cation sublattices form two distinct, intertwined helices, reminiscent of DNA. This topology is stabilized by a collective helical twisting of the oxygen octahedral framework, which gives rise to an electric Dzyaloshinskii-Moriya-like interaction. The resulting structure, which can be canceptualized as a "self-Moir\'e" crystal, exhibits two coupled functionalities. First, it possesses a rotational pseudo-zero-energy mode that underpins a giant piezoelectric response ($e_{33}\approx$16 C/m$^2$). Second, the long-period potential reconstructs the electronic band structure, leading to a multi-valley electronic topology at the valence band edge. Our work establishes a physical route to designing complex chiral order that supports both giant electromechanical coupling and multi-valley electronics.

cond-mat.mtrl-sci↗

arXiv2025

LakeVisage: Towards Scalable, Flexible and Interactive Visualization Recommendation for Data Discovery over Data Lakes

Data discovery from data lakes is an essential application in modern data science. While many previous studies focused on improving the efficiency and effectiveness of data discovery, little attention has been paid to the usability of such applications. In particular, exploring data discovery results can be cumbersome due to the cognitive load involved in understanding raw tabular results and identifying insights to draw conclusions. To address this challenge, we introduce a new problem -- visualization recommendation for data discovery over data lakes -- which aims at automatically identifying visualizations that highlight relevant or desired trends in the results returned by data discovery engines. We propose LakeVisage, an end-to-end framework as the first solution to this problem. Given a data lake, a data discovery engine, and a user-specified query table, LakeVisage intelligently explores the space of visualizations and recommends the most useful and ``interesting'' visualization plans. To this end, we developed (i) approaches to smartly construct the candidate visualization plans from the results of the data discovery engine and (ii) effective pruning strategies to filter out less interesting plans so as to accelerate the visual analysis. Experimental results on real data lakes show that our proposed techniques can lead to an order of magnitude speedup in visualization recommendation. We also conduct a comprehensive user study to demonstrate that LakeVisage offers convenience to users in real data analysis applications by enabling them seamlessly get started with the tasks and performing explorations flexibly.

cs.DB↗

arXiv2024

Reduction-based Pseudo-label Generation for Instance-dependent Partial Label Learning

Instance-dependent Partial Label Learning (ID-PLL) aims to learn a multi-class predictive model given training instances annotated with candidate labels related to features, among which correct labels are hidden fixed but unknown. The previous works involve leveraging the identification capability of the training model itself to iteratively refine supervision information. However, these methods overlook a critical aspect of ID-PLL: the training model is prone to overfitting on incorrect candidate labels, thereby providing poor supervision information and creating a bottleneck in training. In this paper, we propose to leverage reduction-based pseudo-labels to alleviate the influence of incorrect candidate labels and train our predictive model to overcome this bottleneck. Specifically, reduction-based pseudo-labels are generated by performing weighted aggregation on the outputs of a multi-branch auxiliary model, with each branch trained in a label subspace that excludes certain labels. This approach ensures that each branch explicitly avoids the disturbance of the excluded labels, allowing the pseudo-labels provided for instances troubled by these excluded labels to benefit from the unaffected branches. Theoretically, we demonstrate that reduction-based pseudo-labels exhibit greater consistency with the Bayes optimal classifier compared to pseudo-labels directly generated from the predictive model.

cs.LG↗

arXiv2024

Qr-Hint: Actionable Hints Towards Correcting Wrong SQL Queries

We describe a system called Qr-Hint that, given a (correct) target query Q* and a (wrong) working query Q, both expressed in SQL, provides actionable hints for the user to fix the working query so that it becomes semantically equivalent to the target. It is particularly useful in an educational setting, where novices can receive help from Qr-Hint without requiring extensive personal tutoring. Since there are many different ways to write a correct query, we do not want to base our hints completely on how Q* is written; instead, starting with the user's own working query, Qr-Hint purposefully guides the user through a sequence of steps that provably lead to a correct query, which will be equivalent to Q* but may still "look" quite different from it. Ideally, we would like Qr-Hint's hints to lead to the "smallest" possible corrections to Q. However, optimality is not always achievable in this case due to some foundational hurdles such as the undecidability of SQL query equivalence and the complexity of logic minimization. Nonetheless, by carefully decomposing and formulating the problems and developing principled solutions, we are able to provide provably correct and locally optimal hints through Qr-Hint. We show the effectiveness of Qr-Hint through quality and performance experiments as well as a user study in an educational setting.

cs.DB↗

arXiv2024

Giant piezoelectric effects of topological structures in stretched ferroelectric membranes

Freestanding ferroelectric oxide membranes emerge as a promising platform for exploring the interplay between topological polar ordering and dipolar interactions that are continuously tunable by strain. Our investigations combining density functional theory (DFT) and deep-learning-assisted molecular dynamics simulations demonstrate that DFT-predicted strain-driven morphotropic phase boundary involving monoclinic phases manifest as diverse domain structures at room temperatures, featuring continuous distributions of dipole orientations and mobile domain walls. Detailed analysis of dynamic structures reveals that the enhanced piezoelectric response observed in stretched PbTiO$_3$ membranes results from small-angle rotations of dipoles at domain walls, distinct from conventional polarization rotation mechanism and adaptive phase theory inferred from static structures. We identify a ferroelectric topological structure, termed "dipole spiral," which exhibits a giant intrinsic piezoelectric response ($>$320 pC/N). This helical structure, possessing a rotational zero-energy mode, unlocks new possibilities for exploring chiral phonon dynamics and dipolar Dzyaloshinskii-Moriya-like interactions.

cond-mat.mtrl-sci↗

arXiv2023

Depolarization Induced III-V Triatomic Layers with Tristable Polarization States

The integration of ferroelectrics that exhibit high dielectric, piezoelectric, and thermal susceptibilities with the mainstream semiconductor industry will enable novel device types for widespread applications, and yet there are few silicon-compatible ferroelectrics suitable for device downscaling. We demonstrate with first-principles calculations that the enhanced depolarization field at the nanoscale can be utilized to soften unswitchable wurtzite III-V semiconductors, resulting in ultrathin two-dimensional (2D) sheets possessing reversible polarization states. A 2D sheet of AlSb consisting of three atomic planes is identified to host both ferroelectricity and antiferroelectricity, and the tristate switching is accompanied by a metal-semiconductor transition. The thermodynamics stability and potential synthesizability of the triatomic layer are corroborated with phonon spectrum calculations, ab initio molecular dynamics, and variable-composition evolutionary structure search. We propose a 2D AlSb-based homojunction field effect transistor that supports three distinct and nonvolatile resistance states. This new class of III-V semiconductor-derived 2D materials with dual ferroelectricity and antiferroelectricity opens up the possibility for nonvolatile multibit-based integrated nanoelectronics.

cond-mat.mtrl-sci↗

arXiv2022

Origin of ferroelectricity in magnesium doped zinc oxide

Recent experiments demonstrated robust ferroelectricity in Mg-doped ZnO (ZMO) of the wurtzite structure, hinting at a promising strategy to substantially expand the list of ferroelectrics by doping conventional piezoelectrics. We investigate the origin of ferroelectricity in ZMO with first-principles density functional theory (DFT). The general argument that the Mg alloying could soften the ionic potential energy surface of ZMO for polarization reversal is overly simplified. Our DFT calculations reveal that even at a high Mg concentration, the energy difference ($ΔU$) between the polar and nonpolar phases remains prohibitively large for ZMO systems when the strain is fixed to the polar phase. Interestingly, the magnitude of $ΔU$ becomes substantially smaller when the strain relaxation is allowed, approaching the value of typical perovskite ferroelectrics such as PbTiO$_3$ with increasing Mg doping concentrations. The enabled switchability of ZMO systems is attributed to a hexagonal phase of MgO that is much lower in energy than its wurtzite counterpart. Detailed orbital and bonding analysis supports that the intra-atomic $3d_{z^2}$-$4p_z$ orbital self-mixing of Zn plays an important role in stabilizing the polar wurtzite phase, the lack of which is responsible for the low-energy nonpolar hexagonal phase of MgO.

cond-mat.mtrl-sci↗