arXiv subjects

Yuan Lu

Publications and source records attributed to Yuan Lu.

At least 55 records · Page 3Linked to original sources

arXiv2026

GlyphBanana: Advancing Precise Text Rendering Through Agentic Workflows

Despite recent advances in generative models driving significant progress in text rendering, accurately generating complex text and mathematical formulas remains a formidable challenge. This difficulty primarily stems from the limited instruction-following capabilities of current models when encountering out-of-distribution prompts. To address this, we introduce GlyphBanana, alongside a corresponding benchmark specifically designed for rendering complex characters and formulas. GlyphBanana employs an agentic workflow that integrates auxiliary tools to inject glyph templates into both the latent space and attention maps, facilitating the iterative refinement of generated images. Notably, our training-free approach can be seamlessly applied to various Text-to-Image (T2I) models, achieving superior precision compared to existing baselines. Extensive experiments demonstrate the effectiveness of our proposed workflow. Associated code is publicly available at https://github.com/yuriYanZeXuan/GlyphBanana.

cs.CV↗

arXiv2026

Rewards as Labels: Revisiting RLVR from a Classification Perspective

Reinforcement Learning with Verifiable Rewards has recently advanced the capabilities of Large Language Models in complex reasoning tasks by providing explicit rule-based supervision. Among RLVR methods, GRPO and its variants have achieved strong empirical performance. Despite their success, we identify that they suffer from Gradient Misassignment in Positives and Gradient Domination in Negatives, which lead to inefficient and suboptimal policy updates. To address these issues, we propose Rewards as Labels (REAL), a novel framework that revisits verifiable rewards as categorical labels rather than scalar weights, thereby reformulating policy optimization as a classification problem. Building on this, we further introduce anchor logits to enhance policy learning. Our analysis reveals that REAL induces a monotonic and bounded gradient weighting, enabling balanced gradient allocation across rollouts and effectively mitigating the identified mismatches. Extensive experiments on mathematical reasoning benchmarks show that REAL improves training stability and consistently outperforms GRPO and strong variants such as DAPO. On the 1.5B model, REAL improves average Pass@1 over DAPO by 6.7%. These gains further scale to 7B model, REAL continues to outperform DAPO and GSPO by 6.2% and 1.7%, respectively. Notably, even with a vanilla binary cross-entropy, REAL remains stable and exceeds DAPO by 4.5% on average.

cs.LG↗

arXiv2026

Tool-Genesis: A Task-Driven Tool Creation Benchmark for Self-Evolving Language Agent

Research on self-evolving language agents has accelerated, drawing increasing attention to their ability to create, adapt, and maintain tools from task requirements. However, existing benchmarks predominantly rely on predefined specifications, which limits scalability and hinders truly autonomous evolution. While recent studies attempt to dynamically generate tools, they primarily emphasize downstream performance, resulting in a "black-box" evaluation that makes it difficult to attribute failures to specific causes. To address this, we propose Tool-Genesis, a diagnostic benchmark designed to quantify agent capabilities across multiple dimensions, including interface compliance, functional correctness, and downstream utility. Tool-Genesis evaluates whether agents can construct task-relevant tools solely from abstract requirements (without preset specifications) and use them to solve realistic problems. Crucially, we find that even state-of-the-art models struggle to produce precise tool interfaces or executable logic in a one-shot setting. These minor initial flaws are amplified through the pipeline, leading to a sharp degradation in downstream metrics. We hope Tool-Genesis will guide future research toward training and steering models to synthesize persistent, general-purpose tools that better address real-world challenges.

cs.SE↗

arXiv2026

LLaDA2.1: Speeding Up Text Diffusion via Token Editing

While LLaDA2.0 showcased the scaling potential of 100B-level block-diffusion models and their inherent parallelization, the delicate equilibrium between decoding speed and generation quality has remained an elusive frontier. Today, we unveil LLaDA2.1, a paradigm shift designed to transcend this trade-off. By seamlessly weaving Token-to-Token (T2T) editing into the conventional Mask-to-Token (M2T) scheme, we introduce a joint, configurable threshold-decoding scheme. This structural innovation gives rise to two distinct personas: the Speedy Mode (S Mode), which audaciously lowers the M2T threshold to bypass traditional constraints while relying on T2T to refine the output; and the Quality Mode (Q Mode), which leans into conservative thresholds to secure superior benchmark performances with manageable efficiency degrade. Furthering this evolution, underpinned by an expansive context window, we implement the first large-scale Reinforcement Learning (RL) framework specifically tailored for dLLMs, anchored by specialized techniques for stable gradient estimation. This alignment not only sharpens reasoning precision but also elevates instruction-following fidelity, bridging the chasm between diffusion dynamics and complex human intent. We culminate this work by releasing LLaDA2.1-Mini (16B) and LLaDA2.1-Flash (100B). Across 33 rigorous benchmarks, LLaDA2.1 delivers strong task performance and lightning-fast decoding speed. Despite its 100B volume, on coding tasks it attains an astounding 892 TPS on HumanEval+, 801 TPS on BigCodeBench, and 663 TPS on LiveCodeBench.

cs.LG↗

arXiv2026

DeepAgent: A General Reasoning Agent with Scalable Toolsets

Large reasoning models have demonstrated strong problem-solving abilities, yet real-world tasks often require external tools and long-horizon interactions. Existing agent frameworks typically follow predefined workflows, which limit autonomous and global task completion. In this paper, we introduce DeepAgent, an end-to-end deep reasoning agent that performs autonomous thinking, tool discovery, and action execution within a single, coherent reasoning process. To manage long-horizon interactions, we introduce an autonomous memory folding mechanism that compresses past interactions into structured episodic, working, and tool memories, reducing error accumulation while preserving critical information. To teach general-purpose tool use efficiently and stably, we develop an end-to-end reinforcement learning strategy, namely ToolPO, that leverages LLM-simulated APIs and applies tool-call advantage attribution to assign fine-grained credit to the tool invocation tokens. Extensive experiments on eight benchmarks, including general tool-use tasks (ToolBench, API-Bank, TMDB, Spotify, ToolHop) and downstream applications (ALFWorld, WebShop, GAIA, HLE), demonstrate that DeepAgent consistently outperforms baselines across both labeled-tool and open-set tool retrieval scenarios. The code and demo are available at https://github.com/RUC-NLPIR/DeepAgent.

cs.AI↗

arXiv2026

UniPACT: A Multimodal Framework for Prognostic Question Answering on Raw ECG and Structured EHR

Accurate clinical prognosis requires synthesizing structured Electronic Health Records (EHRs) with real-time physiological signals like the Electrocardiogram (ECG). Large Language Models (LLMs) offer a powerful reasoning engine for this task but struggle to natively process these heterogeneous, non-textual data types. To address this, we propose UniPACT (Unified Prognostic Question Answering for Clinical Time-series), a unified framework for prognostic question answering that bridges this modality gap. UniPACT's core contribution is a structured prompting mechanism that converts numerical EHR data into semantically rich text. This textualized patient context is then fused with representations learned directly from raw ECG waveforms, enabling an LLM to reason over both modalities holistically. We evaluate UniPACT on the comprehensive MDS-ED benchmark, it achieves a state-of-the-art mean AUROC of 89.37% across a diverse set of prognostic tasks including diagnosis, deterioration, ICU admission, and mortality, outperforming specialized baselines. Further analysis demonstrates that our multimodal, multi-task approach is critical for performance and provides robustness in missing data scenarios.

cs.LG↗

arXiv2026

TourPlanner: A Competitive Consensus Framework with Constraint-Gated Reinforcement Learning for Travel Planning

Travel planning is a sophisticated decision-making process that requires synthesizing multifaceted information to construct itineraries. However, existing travel planning approaches face several challenges: (1) Pruning candidate points of interest (POIs) while maintaining a high recall rate; (2) A single reasoning path restricts the exploration capability within the feasible solution space for travel planning; (3) Simultaneously optimizing hard constraints and soft constraints remains a significant difficulty. To address these challenges, we propose TourPlanner, a comprehensive framework featuring multi-path reasoning and constraint-gated reinforcement learning. Specifically, we first introduce a Personalized Recall and Spatial Optimization (PReSO) workflow to construct spatially-aware candidate POIs' set. Subsequently, we propose Competitive consensus Chain-of-Thought (CCoT), a multi-path reasoning paradigm that improves the ability of exploring the feasible solution space. To further refine the plan, we integrate a sigmoid-based gating mechanism into the reinforcement learning stage, which dynamically prioritizes soft-constraint satisfaction only after hard constraints are met. Experimental results on travel planning benchmarks demonstrate that TourPlanner achieves state-of-the-art performance, significantly surpassing existing methods in both feasibility and user-preference alignment.

cs.AI↗

arXiv2025

Agent2World: Learning to Generate Symbolic World Models via Adaptive Multi-Agent Feedback

Symbolic world models (e.g., PDDL domains or executable simulators) are central to model-based planning, but training LLMs to generate such world models is limited by the lack of large-scale verifiable supervision. Current approaches rely primarily on static validation methods that fail to catch behavior-level errors arising from interactive execution. In this paper, we propose Agent2World, a tool-augmented multi-agent framework that achieves strong inference-time world-model generation and also serves as a data engine for supervised fine-tuning, by grounding generation in multi-agent feedback. Agent2World follows a three-stage pipeline: (i) A Deep Researcher agent performs knowledge synthesis by web searching to address specification gaps; (ii) A Model Developer agent implements executable world models; And (iii) a specialized Testing Team conducts adaptive unit testing and simulation-based validation. Agent2World demonstrates superior inference-time performance across three benchmarks spanning both Planning Domain Definition Language (PDDL) and executable code representations, achieving consistent state-of-the-art results. Beyond inference, Testing Team serves as an interactive environment for the Model Developer, providing behavior-aware adaptive feedback that yields multi-turn training trajectories. The model fine-tuned on these trajectories substantially improves world-model generation, yielding an average relative gain of 30.95% over the same model before training. Project page: https://agent2world.github.io.

cs.AI↗

arXiv2025

Unified Acoustic Representations for Screening Neurological and Respiratory Pathologies from Voice

Voice-based health assessment offers unprecedented opportunities for scalable, non-invasive disease screening, yet existing approaches typically focus on single conditions and fail to leverage the rich, multi-faceted information embedded in speech. We present MARVEL (Multi-task Acoustic Representations for Voice-based Health Analysis), a privacy-conscious multitask learning framework that simultaneously detects nine distinct neurological, respiratory, and voice disorders using only derived acoustic features, eliminating the need for raw audio transmission. Our dual-branch architecture employs specialized encoders with task-specific heads sharing a common acoustic backbone, enabling effective cross-condition knowledge transfer. Evaluated on the large-scale Bridge2AI-Voice v2.0 dataset, MARVEL achieves an overall AUROC of 0.78, with exceptional performance on neurological disorders (AUROC = 0.89), particularly for Alzheimer's disease/mild cognitive impairment (AUROC = 0.97). Our framework consistently outperforms single-modal baselines by 5-19% and surpasses state-of-the-art self-supervised models on 7 of 9 tasks, while correlation analysis reveals that the learned representations exhibit meaningful similarities with established acoustic features, indicating that the model's internal representations are consistent with clinically recognized acoustic patterns. By demonstrating that a single unified model can effectively screen for diverse conditions, this work establishes a foundation for deployable voice-based diagnostics in resource-constrained and remote healthcare settings.

cs.SD↗

arXiv2025

Superconducting spintronics with electron symmetry filtering and interfacial spin-orbit coupling

Over the recent years, crossroads of magnetism and superconductivity led to the emerging field of superconducting spintronics. A cornerstone of this venture is the generation of equal-spin triplet Cooper pairs in superconductor-ferromagnet hybrids, enabling long-range spin-polarized supercurrents and magnetic control over superconducting quantum states for the development of energy-efficient cryogenic devices. Until now, nearly all superconducting spintronic devices have relied on direct interfaces between superconductors and ferromagnets, since it was believed that an insulating barrier would decouple spin and charge transport. This assumption, however, appears to be invalid when a thin spin- and orbit-filtering barrier couples epitaxial ferromagnet and the superconductor. Symmetry filtering plays a crucial role in enhancing giant tunneling magnetoresistance (TMR) by selectively allowing specific electronic states to tunnel through the barrier. Such a mechanism is key for high-performance spintronic devices like magnetic random access memories, magnetic sensors or spin-light emitting diodes. This manuscript provides a comprehensive review of superconducting spintronics driven by electron symmetry filtering and interfacial SOC. It emphasizes the critical role of a crystalline MgO barrier in selectively transmitting specific electronic states between V(100) and Fe(100). The manuscript also highlights how interfacial SOC enables symmetry mixing, allowing for the interaction between ferromagnetic and superconducting orderings though MgO(100). This mutual interaction, mediated by interfacial SOC, facilitates the conversion of spin-singlet to spin-triplet Cooper pairs. The work provides key insights into designing SOC based superconductor-ferromagnet hybrid structures for advanced superconducting spintronic functionalities.

cond-mat.supr-con↗

arXiv2025

Localized Conformal Multi-Quantile Regression

Standard conformal prediction methods guarantee marginal coverage but often produce inefficient intervals that fail to adapt to local heteroscedasticity, while recent localized approaches often struggle to maintain validity across distinct subpopulations with varying noise profiles. To address these challenges, we introduce Localized Conformal Multi-Quantile Regression (LCMQR), a novel framework that synergizes multi-quantile information with kernel-based localization to construct efficient and adaptive prediction intervals. Theoretically, we resolve an inconsistency in Conformalized Composite Quantile Regression (CCQR) by proving that our consistent Average-then-Max scoring mechanism systematically yields tighter intervals than the Max-then-Average approach used in prior work. For heterogeneous environments, we extend this framework to Group-Calibrated LCMQR (GC-LCMQR) via a stratified calibration step that guarantees finite-sample validity within distinct subgroups. Experiments on benchmark datasets and an Individual Treatment Effect (ITE) task demonstrate that LCMQR achieves superior efficiency on standard benchmarks, while GC-LCMQR uniquely achieves group-level coverage for target subgroups in mixture populations where baselines fail.

stat.ME↗

arXiv2025

Synthetic Curriculum Reinforces Compositional Text-to-Image Generation

Text-to-Image (T2I) generation has long been an open problem, with compositional synthesis remaining particularly challenging. This task requires accurate rendering of complex scenes containing multiple objects that exhibit diverse attributes as well as intricate spatial and semantic relationships, demanding both precise object placement and coherent inter-object interactions. In this paper, we propose a novel compositional curriculum reinforcement learning framework named CompGen that addresses compositional weakness in existing T2I models. Specifically, we leverage scene graphs to establish a novel difficulty criterion for compositional ability and develop a corresponding adaptive Markov Chain Monte Carlo graph sampling algorithm. This difficulty-aware approach enables the synthesis of training curriculum data that progressively optimize T2I models through reinforcement learning. We integrate our curriculum learning approach into Group Relative Policy Optimization (GRPO) and investigate different curriculum scheduling strategies. Our experiments reveal that CompGen exhibits distinct scaling curves under different curriculum scheduling strategies, with easy-to-hard and Gaussian sampling strategies yielding superior scaling performance compared to random sampling. Extensive experiments demonstrate that CompGen significantly enhances compositional generation capabilities for both diffusion-based and auto-regressive T2I models, highlighting its effectiveness in improving the compositional T2I generation systems.

cs.CV↗

arXiv2025

On O'Grady's generalized Franchetta conjecture for genus 11 K3 surfaces

O'Grady's generalized Franchetta conjecture asks whether any codimension two cycle on the universal polarized K3 surface restricts to a multiple of the Beauville--Voisin class on a given K3 surface. We apply Mukai's program for genus 11 curves and K3 surfaces, together with a result on the tautological generation of the second Chow group of the moduli space of curves, to give an affirmative answer to this conjecture in genus 11.

math.AG↗

arXiv2025

LoopTool: Closing the Data-Training Loop for Robust LLM Tool Calls

Augmenting Large Language Models (LLMs) with external tools enables them to execute complex, multi-step tasks. However, tool learning is hampered by the static synthetic data pipelines where data generation and model training are executed as two separate, non-interactive processes. This approach fails to adaptively focus on a model's specific weaknesses and allows noisy labels to persist, degrading training efficiency. We introduce LoopTool, a fully automated, model-aware data evolution framework that closes this loop by tightly integrating data synthesis and model training. LoopTool iteratively refines both the data and the model through three synergistic modules: (1) Greedy Capability Probing (GCP) diagnoses the model's mastered and failed capabilities; (2) Judgement-Guided Label Verification (JGLV) uses an open-source judge model to find and correct annotation errors, progressively purifying the dataset; and (3) Error-Driven Data Expansion (EDDE) generates new, challenging samples based on identified failures. This closed-loop process operates within a cost-effective, open-source ecosystem, eliminating dependence on expensive closed-source APIs. Experiments show that our 8B model trained with LoopTool significantly surpasses its 32B data generator and achieves new state-of-the-art results on the BFCL-v3 and ACEBench benchmarks for its scale. Our work demonstrates that closed-loop, self-refining data pipelines can dramatically enhance the tool-use capabilities of LLMs.

cs.CL↗

arXiv2025

HyCTAS: Multi-Objective Hybrid Convolution-Transformer Architecture Search for Real-Time Image Segmentation

Real-time image segmentation demands architectures that preserve fine spatial detail while capturing global context under tight latency and memory budgets. Image segmentation is one of the most fundamental problems in computer vision and has drawn a lot of attention due to its vast applications in image understanding and autonomous driving. However, designing effective and efficient segmentation neural architectures is a labor-intensive process that may require numerous trials by human experts. In this paper, we address the challenge of integrating multi-head self-attention into high-resolution representation CNNs efficiently by leveraging architecture search. Manually replacing convolution layers with multi-head self-attention is non-trivial due to the costly overhead in memory to maintain high resolution. By contrast, we develop a multi-target multi-branch supernet method, which not only fully utilizes the advantages of high-resolution features but also finds the proper location for placing the multi-head self-attention module. Our search algorithm is optimized towards multiple objectives (e.g., latency and mIoU) and is capable of finding architectures on the approximate Pareto front with an arbitrary number of branches in a single search. We further present a series of models via the Hybrid Convolutional-Transformer Architecture Search (HyCTAS) method that searches for the best hybrid combination of lightweight convolution layers and memory-efficient self-attention layers between branches from different resolutions and fuses them at high resolution for both efficiency and effectiveness. On Cityscapes, ADE20K, and COCO, HyCTAS discovers competitive real-time models without ImageNet pretraining, delivering strong accuracy and latency trade-offs. Code and models are available at https://github.com/MarvinYu1995/HyCTAS.

cs.CV↗

arXiv2025

Evolve to Inspire: Novelty Search for Diverse Image Generation

Text-to-image diffusion models, while proficient at generating high-fidelity images, often suffer from limited output diversity, hindering their application in exploratory and ideation tasks. Existing prompt optimization techniques typically target aesthetic fitness or are ill-suited to the creative visual domain. To address this shortcoming, we introduce WANDER, a novelty search-based approach to generating diverse sets of images from a single input prompt. WANDER operates directly on natural language prompts, employing a Large Language Model (LLM) for semantic evolution of diverse sets of images, and using CLIP embeddings to quantify novelty. We additionally apply emitters to guide the search into distinct regions of the prompt space, and demonstrate that they boost the diversity of the generated images. Empirical evaluations using FLUX-DEV for generation and GPT-4o-mini for mutation demonstrate that WANDER significantly outperforms existing evolutionary prompt optimization baselines in diversity metrics. Ablation studies confirm the efficacy of emitters.

cs.CV↗

arXiv2025

Fints: Efficient Inference-Time Personalization for LLMs with Fine-Grained Instance-Tailored Steering

The rapid evolution of large language models (LLMs) has intensified the demand for effective personalization techniques that can adapt model behavior to individual user preferences. Despite the non-parametric methods utilizing the in-context learning ability of LLMs, recent parametric adaptation methods, including personalized parameter-efficient fine-tuning and reward modeling emerge. However, these methods face limitations in handling dynamic user patterns and high data sparsity scenarios, due to low adaptability and data efficiency. To address these challenges, we propose a fine-grained and instance-tailored steering framework that dynamically generates sample-level interference vectors from user data and injects them into the model's forward pass for personalized adaptation. Our approach introduces two key technical innovations: a fine-grained steering component that captures nuanced signals by hooking activations from attention and MLP layers, and an input-aware aggregation module that synthesizes these signals into contextually relevant enhancements. The method demonstrates high flexibility and data efficiency, excelling in fast-changing distribution and high data sparsity scenarios. In addition, the proposed method is orthogonal to existing methods and operates as a plug-in component compatible with different personalization techniques. Extensive experiments across diverse scenarios--including short-to-long text generation, and web function calling--validate the effectiveness and compatibility of our approach. Results show that our method significantly enhances personalization performance in fast-shifting environments while maintaining robustness across varying interaction modes and context lengths. Implementation is available at https://github.com/KounianhuaDu/Fints.

cs.AI↗

arXiv2025

Video-Thinker: Sparking "Thinking with Videos" via Reinforcement Learning

Recent advances in image reasoning methods, particularly "Thinking with Images", have demonstrated remarkable success in Multimodal Large Language Models (MLLMs); however, this dynamic reasoning paradigm has not yet been extended to video reasoning tasks. In this paper, we propose Video-Thinker, which empowers MLLMs to think with videos by autonomously leveraging their intrinsic "grounding" and "captioning" capabilities to generate reasoning clues throughout the inference process. To spark this capability, we construct Video-Thinker-10K, a curated dataset featuring autonomous tool usage within chain-of-thought reasoning sequences. Our training strategy begins with Supervised Fine-Tuning (SFT) to learn the reasoning format, followed by Group Relative Policy Optimization (GRPO) to strengthen this reasoning capability. Through this approach, Video-Thinker enables MLLMs to autonomously navigate grounding and captioning tasks for video reasoning, eliminating the need for constructing and calling external tools. Extensive experiments demonstrate that Video-Thinker achieves significant performance gains on both in-domain tasks and challenging out-of-domain video reasoning benchmarks, including Video-Holmes, CG-Bench-Reasoning, and VRBench. Our Video-Thinker-7B substantially outperforms existing baselines such as Video-R1 and establishes state-of-the-art performance among 7B-sized MLLMs.

cs.CV↗