Searcharxiv⌕ Search

arXiv subjects

Heng Li

Publications and source records attributed to Heng Li.

At least 55 records · Page 3Linked to original sources
arXiv2026

PTCOG Treatment Efficiency Subcommittee Risk Assessment Report on Patient-Specific Quality Assurance

Patient-specific quality assurance (PSQA) in pencil beam scanning proton therapy (PBS-PT) is often treated as a purely technical verification task. This PTCOG Treatment Efficiency Subcommittee White Paper instead frames PSQA as a workflow-embedded risk-control strategy and asks how different PSQA approaches reshape the same clinical risk landscape. Using a generic PBS-PT process-driven Failure Mode and Effects Analysis (pFMEA), 44 validated PSQA-relevant failure modes across 20 process steps were scored under a common no-PSQA baseline and three PSQA pathways: measurement-based PSQA, log file-based PSQA, and independent secondary dose calculation. A staged mathematical formalism separates preparatory data-stage effects, method-specific full-stage verification, cumulative endstate effects, and a Data-to-Cum bridge that quantifies additional verification benefit on the baseline scale. In this expert-scored, baseline-anchored model, log file-based PSQA produced the largest cumulative workflow-level risk-score reduction, followed by measurement-based PSQA and independent secondary dose calculation. The ranking is not a winner-takes-all rule or probability-calibrated risk estimate; instead, each method shows distinct risk-control strengths in different workflow regions. The White Paper therefore supports a risk-informed hybrid PSQA architecture, where log file-based PSQA, measurement-based PSQA, and independent secondary dose calculation are assigned to the workflow segments in which their signatures are strongest. It provides a transparent, semi-quantitative, stage-resolved framework for institutions seeking to evaluate, implement, or evolve PSQA in PBS-PT and emphasizes that log file-based PSQA must itself be supported by validated and governed log data and treatment records.

physics.med-ph↗
arXiv2026

Strong counterexamples to Mubayi's supersaturation conjecture in every uniformity

The supersaturation problem asks, for a fixed $r$-graph $\mathcal F$, for the minimum number of copies of $\mathcal F$ in an $n$-vertex $r$-graph with $\ex(n,\mathcal F)+q$ edges. Mubayi conjectured a local form of supersaturation under a stability hypothesis: if $\mathcal F$ is non-$r$-partite and stable, meaning roughly that the extremal $\mathcal F$-free construction is unique and all near-extremal $\mathcal F$-free $r$-graphs are close to it, then this minimum should be at least $q c(n,\mathcal F)$, where $c(n,\mathcal F)$ is the minimum number of copies created by adding one edge to the extremal $\mathcal F$-free $r$-graph. We disprove this conjectured local lower bound in every uniformity. For every $r\ge2$ and every $K>1$, we construct a stable $r$-graph $\mathcal F$ such that, for all sufficiently large $n$ and every $1\le q\le δn$, there is an $n$-vertex $r$-graph with $\ex(n,\mathcal F)+q$ edges and at most $K^{-1}q c(n,\mathcal F)$ copies of $\mathcal F$. Thus the conjectured lower bound can already fail at $q=1$, and the failure can be by an arbitrarily large constant factor in every uniformity.

math.CO↗
arXiv2026

ViPSim: Collaborating Visual and Parameter Spaces for Consistent Long-Horizon Embodied World Models

Embodied World Models (EWMs) have emerged as a scalable and risk-free paradigm for advancing embodied intelligence, enabling the safety-critical evaluation of Vision-Language-Action systems. However, their reliability as evaluation benchmarks and foundational simulators is often hindered by the representation gap between low-dimensional actions and high-dimensional video synthesis. This gap results in a lack of geometric correspondence, manifesting as accumulated trajectory drift and inconsistent robot-object interactions during long-horizon rollouts. To bridge this gap, we propose ViPSim, a framework that achieves consistent long-horizon generation through the synergistic collaboration of Visual and Parameter Spaces. We define the Visual Space as a domain of explicit spatial priors, integrating pixel-aligned projections of end-effector pose, camera perspectives, depth-informed scene geometry, and robotic morphological masks to provide dense structural grounding. Concurrently, the Parameter Space serves as a domain of numerical drivers, injecting raw action sequences and camera matrices to provide precise motion guidance. By unifying these two spaces, ViPSim ensures that the generated states are simultaneously anchored by geometric boundaries and steered by numerical commands. Extensive experiments demonstrate that ViPSim is backbone-agnostic and significantly enhances trajectory consistency. Notably, our approach exhibits emergent capabilities in generating complex interactions with deformable objects (e.g., cloth folding) and maintains robust performance in out-of-distribution and cross-embodiment scenarios, providing a high-fidelity foundation for the automated evaluation and predictive control of embodied agents.

cs.CV↗
arXiv2026

On Mubayi's Polynomial-Ideal Conjecture and a Hypergraph Turán Theorem

Among the many proofs of Turán's classical theorem, one particularly surprising proof due to Li and Li uses ideals in polynomial rings to record missing edges. Motivated by their proof, Mubayi proposed a hypergraph analogue, conjecturing that an ideal generated by multipartite $3$-graphs coincides with a differentiated diagonal-vanishing ideal. If true, this conjecture would imply the extremal-number part of Mubayi's classical hypergraph extension of Turán's theorem. We disprove this conjecture throughout the nontrivial parameter range. We then give an alternative algebraic proof of Mubayi's extremal formula for the family $\mathcal{K}_{\ell}^{(r)}$ of clique expansions, using monomial cover ideals and a Hilbert-function symmetrization theorem for square-zero quadratic monomial quotients.

math.CO↗
arXiv2026

Sharp bounds for minimal dependencies of linear-form powers

Motivated by the dimension-bound part of a problem of Bukh, we study Veronese circuits: how large can the span of $t$ linear forms be if their $m$-th powers are minimally linearly dependent? We prove the sharp finite dimension bound \[ \dim L\leq \frac{t+m-2}{m}. \] Here $\ell_1,\ldots,\ell_t$ are nonzero homogeneous linear forms over a field of characteristic zero, the powers $\ell_1^m,\ldots,\ell_t^m$ form a circuit, and $L=\Span\{\ell_1,\ldots,\ell_t\}$. Rational-normal-curve configurations attain equality for infinitely many pairs $(t,\dim L)$; in particular, the affine bound itself is sharp and the optimal leading constant in Bukh's question is $1/m$. The proof uses a coding-theoretic translation: the coefficient row space of the powers is the $m$-th Schur power of the coefficient code, and the minimality hypothesis makes this Schur power a full-support hyperplane to which the Schur-product Kneser theorem of Mirandola and Zémor applies. The same method yields flat-concentration and interpolation criteria, a Cayley--Bacharach lower bound, Segre--Veronese and positive-characteristic variants, and Hilbert-function constraints for equality and near-equality in Veronese circuits.

math.CO↗
arXiv2026

Fast genomic read alignment with minibwa

Motivation: BWA-MEM remains a popular short-read mapper especially for the purpose of variant calling. Several groups have accelerated this algorithm as it has been the performance bottleneck of many current workflows. However, constrained by the original design, these drop-in replacements could only achieve limited speedup. Breaking changes to BWA-MEM are required for further improvement. Results: We developed minibwa for aligning short and accurate long reads against a reference genome. It combines BWA-MEM variable-length seeding with minimap2 chaining and base alignment. It speeds up BWA-MEM2 further with additional prefetch for seeding, new heuristics to skip unnecessary mate rescue and reduced effort in highly repetitive regions where reads would anyway be wrongly mapped due to structural changes. Minibwa is about four times as fast as BWA-MEM and over twice as fast as BWA-MEM2 at comparable accuracy. It also natively supports directional bisulfite sequencing data to high mapping accuracy. Availability and implementation: https://github.com/lh3/minibwa

q-bio.GN↗
arXiv2026

Sentinel: Decoding Context Utilization via Attention Probing for Efficient LLM Context Compression

Retrieval-augmented generation (RAG) often suffers from long and noisy retrieved contexts. Existing context compression methods typically rely on heuristic relevance estimation or supervised compression models rather than on how LLMs utilize retrieved context during inference. We propose Sentinel, a lightweight sentence-level compression framework that decodes inference-time contextual utilization behaviors from head-wise attention patterns of frozen LLMs. To ground supervision in retrieval-dependent answering behavior, Sentinel trains a lightweight probe using QA examples where the model succeeds only when retrieved context is available. Sentinel performs compression using only a single non-autoregressive forward pass without dedicated compression training or autoregressive scoring. Empirically, we find that effective contextual utilization signals remain accessible even in compact proxy models. On LongBench, Sentinel with a 0.5B proxy model achieves up to 5$\times$ compression while attaining question-answering performance competitive with compression methods built on 7B-scale models. Despite being trained only on English QA data, Sentinel also generalizes effectively to Chinese and out-of-domain settings.

cs.CL↗
arXiv2026

MaskWAM: Unifying Mask Prompting and Prediction for World-Action Models

World Action Models (WAMs) present a promising paradigm for robotic control via video prediction. However, current WAMs suffer from fundamental spatial bottlenecks: standard text inputs introduce referential ambiguity in cluttered scenes, while unstructured RGB predictions lack semantic grounding and remain biased by task-irrelevant backgrounds. To overcome these limitations, we introduce MaskWAM, an object-centric world-action model. By jointly integrating masks as both explicit inputs and predictions via a unified Mixture of Transformers (MoT), MaskWAM unlocks robust policy generalization. This design provides two key benefits: (1) predicting future masks yields object-centric semantic supervision that suppresses visual noise, significantly enhancing even standard text-conditioned WAMs; and (2) coupling this predictive supervision with first-frame visual prompts, such as target object masks, establishes a precise spatial anchor that substantially reduces language ambiguity. Crucially, as WAMs are inherently vision-driven architectures, direct mask conditioning yields substantially stronger guidance than text alone, establishing a precise and robust paradigm for manipulating unseen objects. Evaluations on LIBERO, RoboTwin, and real-world tasks demonstrate that MaskWAM significantly outperforms baselines in both language-clear and language-ambiguous tasks.

cs.CV↗
arXiv2026

On a hypergraph Turán problem of Balogh-Bohman-Bollobás-Zhao

Let $S$ and $T$ be disjoint sets with $|S|=i$ and $|T|=r-1$ for $2\le i\le r-1$, and let $B_i^{(r)}$ be the $r$-graph on $S\cup T$ whose edges are the $r$-subsets containing $S$ or $T$. We study the deficit $q_{r,i}:=1-π(B_i^{(r)})$ in its Turán density. Balogh, Bohman, Bollobás, and Zhao previously obtained bounds for these deficits with logarithmic gaps near both ends of the sequence $B_i^{(r)}$, namely, when $i=O(1)$ or $i=r-O(1)$. We close these gaps by showing that, as $r\to\infty$, for every fixed integer $a\ge1$, $q_{r,a+1}=Θ_a(r^{-a})$, and for every fixed integer $b\ge2$, $q_{r,r-b}=Θ_b(r^{-b}\log r)$.

math.CO↗
arXiv2026

Physical-Space Scarring in Generic Bunimovich Stadia

For the family of Dirichlet stadia $S_t$ whose rectangular part has height $π$ and half-length $πt/2$, $t \in [1,2]$, we show that for Lebesgue almost every $t$ there exist real eigenfunctions $u_j$ and a smooth mean-zero physical observable $Q$ for which $\langle Q u_j,u_j\rangle$ has a non-zero subsequential limit. Consequently, along the same subsequence, the eigenfunction mass fails to equidistribute on a fixed region whose relative boundary in the interior of the stadium is smooth. This proves a physical-space strengthening of Hassell's non-QUE theorem for generic stadia, and thus gives an affirmative answer to Tao's question in Hassell's generic setting. The proof uses the classification of generic stadia in Hassell's argument. In each of the resulting cases, we construct an appropriate physical observable $Q$ that converts Hassell's phase-space obstruction to QUE into physical-space non-equidistribution.

math.AP↗
arXiv2026

An Empirical Study on Logging Evolution On Stack Overflow: Trends, Topics, and Challenges

Context: Logging is a crucial practice in software engineering, aiding developers in debugging applications when errors occur. While existing research has explored logging challenges from an academic perspective through literature reviews and source code analysis, a comprehensive study from the practitioners' perspective remains lacking. Objective: This paper aims to bridge this knowledge gap by presenting an in-depth analysis of trends, topics, and challenges in logging based on a dataset of 216,094 posts from Stack Overflow (SO), a popular Q\&A platform for developers. Method: We analyzed longitudinal trends by examining metadata related to users, questions, and tags associated with logging discussions. To identify prevalent discussion topics, we employed a Large Language Model (LLM)--based classification approach, based on a manually validated ground-truth sample. Topic popularity was assessed through average scores and views, while difficulty was measured using three community-driven metrics: the proportion of questions without accepted answers, the proportion of unanswered questions, and the median time to receive an accepted answer. Results: Our analysis identifies 11 distinct topics, with the top three (General Logging Practices, Error Handling and Debugging, and Logging Levels and Output) accounting for over 70\% of all logging-related discussions. Notably, Logging in Containerized Environments emerged as the most difficult topic: 64.9\% of its questions lack an accepted answer, and its median resolution time is among the highest. These findings highlight enduring practitioner struggles with logging in Docker or other containerized environments and the integration of logging pipelines into orchestrators such as Kubernetes and cloud environments. Conclusion: This study sheds light on the practical challenges of logging and provides actionable insights for developers, framework vendors, researchers, and educators.

cs.SE↗
arXiv2026

A single $3$-graph with infinite stability number

The stability number of a forbidden family measures how many different structures are needed to approximate all near-extremal constructions avoiding it. An infinite stability number means that no finite list of structures suffices. We construct a simple explicit $3$-graph whose stability number is infinite. This extends the infinite-stability phenomenon for finite forbidden families, established by Hou--Li--Liu--Mubayi--Zhang, to the single-forbidden setting, and further develops the single-$3$-graph direction of Balogh--Clemen--Luo, in which exponentially many exact extremal constructions coexist with stability.

math.CO↗
arXiv2026

SeePhys Pro: Diagnosing Modality Transfer and Blind-Training Effects in Multimodal RLVR for Physics Reasoning

We introduce SeePhys Pro, a fine-grained modality transfer benchmark that studies whether models preserve the same reasoning capability when critical information is progressively transferred from text to image. Unlike standard vision-essential benchmarks that evaluate a single input form, SeePhys Pro features four semantically aligned variants for each problem with progressively increasing visual elements. Our evaluation shows that current frontier models are far from representation-invariant reasoners: performance degrades on average as information moves from language to diagrams, with visual variable grounding as the most critical bottleneck. Motivated by this inference-time fragility, we further develop large training corpora for multimodal RLVR and use blind training as a diagnostic control, finding that RL with all training images masked can still improve performance on unmasked validation sets. To analyze this effect, text-deletion, image-mask-rate, and format-saturation controls suggest that such gains can arise from residual textual and distributional cues rather than valid visual evidence. Our results highlight the need to evaluate multimodal reasoning not only by final-answer accuracy, but also by robustness under modality transfer and by diagnostics that test whether improvements rely on task-critical visual evidence.

cs.AI↗
arXiv2026

Continuous Expert Assembly: Instance-Conditioned Low-Rank Residuals for All-in-One Image Restoration

Real-world image degradation is often unknown, spatially non-uniform, and compositional, requiring all-in-one restoration models to adapt a single set of weights to diverse local corruption patterns without test-time degradation labels. Existing methods typically modulate a shared backbone with global prompts or degradation descriptors, or route features through predefined expert pools. However, compact global conditioning can bottleneck localized degradation evidence, while static expert routing may produce homogeneous updates or rely on unstable sparse assignments. We propose \textbf{Continuous Expert Assembly} (CEA), a token-wise dynamic parameterization framework for all-in-one image restoration. CEA employs a lightweight \textbf{Cross-Attention Hyper-Adapter} to probe intermediate spatial features and synthesize instance-conditioned low-rank routing bases and residual directions. Each spatial token then assembles its own residual update via dense signed dot-product affinities over the generated rank-wise components, avoiding external prompts, static expert banks, and discrete Top- selection. The resulting assembly rule also admits a linear-attention perspective, making its dense token-wise routing behavior transparent. Experiments on AIO-3, AIO-5, and CDD-11 show that CEA improves average restoration quality over strong prompt-, descriptor-, and expert-based baselines, with the clearest gains on spatially varying and compositional degradations, while maintaining favorable parameter, FLOP, and runtime efficiency.

cs.CV↗
arXiv2026

MotionHiFlow: Text-to-motion via hierarchical flow matching

Text-to-motion generation aims to generate 3D human motions that are tightly aligned with the input text while remaining physically plausible and rich in fine-grained detail. Although recent approaches can produce complex and natural movements, they usually operate at only one temporal scale, which limits both semantic alignment and temporal coherence. Inspired by the fact that complex motions are conceptualized hierarchically rather than at a single temporal scale in the human cognitive system, we propose \textit{MotionHiFlow}, a hierarchical flow matching framework to generate motion progressively by constructing flow path from low to high temporal scales. The flows at lower scales capture high-level semantics and coarse motion structures, while flows at higher scales refine temporal details. To link the flows across scales, we introduce a novel cross-scale transition process, ensuring continuity and preserving noise consistency. Furthermore, by integrating a Text-Motion Diffusion Transformer and a topology-aware Motion VAE, MotionHiFlow explicitly models structural dependencies among joints via joint-aware positional encoding and skeletal topology, enabling precise semantic alignment alongside fine-grained motion details. Extensive experiments on HumanML3D and KIT-ML benchmarks demonstrate state-of-the-art performance, with ablation studies confirming the effectiveness of the hierarchical design and key components. Code is available at https://github.com/ai-lh/MotionHiFlow.

cs.CV↗
arXiv2026

Ethics Testing: Proactive Identification of Generative AI System Harms

Generative Artificial Intelligence (GAI) systems that can automatically generate content in the form of source code or other contents (e.g., images) has seen increasing popularity due to the emergence of tools such as ChatGPT which rely on Large Language Models (LLMs). Misuse of the automatically generated content can incur serious consequences due to potential harms in the generated content. Despite the importance of ensuring the quality of automatically generated content, there is little to no approach that can systematically generate tests for identifying software harms in the content generated by these GAI systems. In this article, we introduce the novel concept of ethics testing which aims to systematically generate tests for identifying software harms. Different from existing testing methodologies (e.g., fairness testing that aims to identifying software discrimination), ethics testing aims to systematically detect software harms that could be induced due to unethical behavior (e.g., harmful behavior or behavior that violates intellectual property rights) in automatically generated content. We introduced the concept of ethics testing, discussed the challenges therewithin, and conducted five case studies to show how ethics testing can be performed for generative AI systems.

cs.SE↗
arXiv2026

Structural Anchors and Reasoning Fragility:Understanding CoT Robustness in LLM4Code

Chain-of-Thought (CoT) prompting is widely used to elicit explicit reasoning from large language models for code (LLM4Code). However, its impact on robustness and the stability of reasoning trajectories under realistic input perturbations remains poorly understood. Prior work has largely evaluated CoT through final correctness, leaving a critical gap in understanding how CoT reshapes internal uncertainty dynamics and why it sometimes harms rather than helps code generation. We suggest that CoT is not uniformly beneficial; instead, its robustness depends on whether perturbations destabilize structurally sensitive commitment points along the reasoning-to-code trajectory. We conduct a controlled, large-scale empirical study of CoT across six models and two code benchmarks (MHPP and BigCodeBench), subjecting task docstrings to systematic character-, word-, and sentence-level perturbations. We instrument full generation traces with token-level uncertainty and define three novel structural anchors: reasoning-code transition, symbolic commitment, and algorithmic articulation. Findings: (1) CoT does not yield uniform performance or robustness gains: its benefits are contingent on model family, task structure, and prompt explicitness. (2) CoT and No-CoT exhibit distinct robustness profiles, with different perturbation families triggering different failure modes. (3) We identify three recurrent trajectory deformations--Lengthening, Branching, and Simplification--that systematically emerge when perturbations interact with structural anchors and explain failure patterns. (4) Early-stage uncertainty serves as a reliable diagnostic signal for localizing where trajectory instability begins around sensitive anchors. These results provide a unified explanation for CoT's mixed performance and suggest design principles for building more robust reasoning-based code generators.

cs.SE↗
arXiv2026

Transversal Hamilton cycles in digraph collections

Given a collection $\mathcal{D} =\{D_1,D_2,\ldots,D_m\}$ of digraphs on the common vertex set $V$, an $m$-edge digraph $H$ with vertices in $V$ is \textit{transversal} in $\mathcal{D}$ if there exists a bijection $φ:E(H)\rightarrow [m]$ such that $e \in E(D_{φ(e)})$ for all $e\in E(H)$. Ghouila-Houri proved that any $n$-vertex digraph with minimum semi-degree at least $\frac{n}{2}$ contains a directed Hamilton cycle. In this paper, we provide a transversal generalization of Ghouila-Houri's theorem, thereby solving a problem proposed by Chakraborti, Kim, Lee and Seo. Our proof utilizes the absorption method for transversals, the regularity method for digraph collections, as well as the transversal blow-up lemma and the related machinery. As an application, when $n$ is sufficiently large, our result implies the transversal version of Dirac's theorem, which was proved by Joos and Kim.

math.CO↗