Search arXiv⌕ Search

arXiv subjects

Min Huang

Publications and source records attributed to Min Huang.

At least 19 recordsLinked to original sources
arXiv2026

Noncommutative marked surfaces II: tagged triangulations, clusters, and their symmetries

The aim of the paper is to define noncommutative cluster structure on several algebras ${\mathcal A}$ related to marked surfaces possibly with orbifold points of various orders, which includes noncommutative clusters, i.e., embeddings of a given group $G$ into the multiplicative monoid ${\mathcal A}^\times$ and an action of a certain braid-like group $Br_{\mathcal A}$ by automorphisms of each cluster group in a compatible way. For punctured surfaces we construct new symmetries, noncommutative tagged clusters and establish a noncommutative Laurent Phenomenon.

math.RT↗
arXiv2026

SemanticFold: Latent Sequence Compression SeparatesLanguage Modeling, Decodability, and Reasoning

We study whether latent sequence compression of prompt prefixes preserves the capabilities that large language models rely on during inference. We introduce SemanticFold, a compression scheme that folds prefix hidden states at learned boundaries, and evaluate it across five model scales: Qwen3-1.7B, Qwen3-8B, SmolLM2-1.7B, Pythia-1.4B, and Pythia-6.9B. We use a fixed-target protocol: a frozen prefix is executed natively or compressed, and both arms teacher-force identical continuation tokens. This design rules out target-selection explanations for likelihood changes. We examine five endpoint families: fixed-target negative log-likelihood, finite-label reasoning accuracy, linear probe accessibility, open-ended generation, and systems-level memory and latency. We find that compression moves these endpoints non-monotonically and that they do not share a single compression threshold. On Qwen3-1.7B at compression ratio R=1.7, compressed-minus-native mean NLL decreases by 0.135 under paired bootstrap with 10000 draws. On SmolLM2 at R=1.2, the mean change is 0.013 higher than native. On both Pythia checkpoints, NLL is effectively unchanged. An NLL decomposition separating sequence shortening from the learned residual transform shows that the favorable Qwen likelihood is attributable primarily to residual adaptation rather than to shortening alone. MLP-only, which applies the transform without shortening, achieves 0.082 lower NLL than Full SemanticFold. Linear probe accuracy and macro AUC change by less than 0.03 in absolute value across conditions, with confidence intervals crossing zero. We conclude that preservation under latent compression has no single scalar certificate: language-model fit, decodability, and reasoning behavior answer different questions and can move in different directions under the same compression operation.

cs.LG↗
arXiv2026

Generalized quantum cluster structure of reduced Lê--Roger--Yang skein algebras and their Frobenius maps

We establish a generalized quantum cluster algebra structure, defined by Bai--Chen--Ding--Xu, on the reduced Lê--Roger--Yang skein algebra of every quasi-triangulable punctured surface. We define Frobenius maps for generalized quantum cluster algebras at roots of unity and use these maps, together with the cluster algebra structure, to construct Frobenius maps for the (reduced) Lê--Roger--Yang skein algebras of all punctured surfaces. As applications of the Frobenius map, we determine the centers of both the Muller--Roger--Yang skein algebras and the reduced Lê--Roger--Yang skein algebras at roots of unity, and compute their ranks as modules over their centers.

math.QA↗
arXiv2026

GeoRefer-Bench: A Benchmark from Referring Pixels to Verifiable Geospatial Reasoning

Referring segmentation in overhead imagery is inherently relational: a query may ask for the buildings north of the road or the pond closest to a residential area, so the correct referent can contain one object, several objects, or none. Existing benchmarks mainly score mask overlap, which cannot verify whether a model actually resolved the stated spatial relation. We introduce GeoRefer-Bench, a benchmark for verifiable geospatial referring segmentation. Each query is represented by an executable logical form over a metric scene graph, and predictions are evaluated with Exact Query Success (EQS), which is satisfied only when the returned instance set exactly matches the set denoted by the query. GeoRefer-Bench contains 700 whole 2048x2048 UAV scenes (2.94 Gpx) at 12.5 and 25 cm ground sampling distance, 26,217 instances, 142,796 spatial relations, and 20,916 executable queries spanning five reasoning levels. It further includes three paraphrases per query, 24.0% unanswerable queries, 2,477 counterfactual pairs, and five leakage-controlled evaluation splits. An independent audit re-derives object geometry, mask ownership, relation values, query execution, and split provenance, finding zero issues across all 700 scenes. Relation-blind strategies can retain non-trivial mIoU while achieving at most 22.7 EQS overall, showing that overlap alone does not certify relational grounding. Across fifteen current models, the strongest reaches 74.1 EQS but drops from 98.9 at level 1 to 60.5 at level 5, while ten models score below 5 EQS on two-hop queries. GeoRefer-Bench turns geospatial referring segmentation from mask matching into verifiable reference resolution.

cs.CV↗
arXiv2026

ASA: Backbone-Training-Free Representation Engineering for Tool-Calling Agents

Adapting LLM agents to domain-specific tool calling remains notably brittle under evolving interfaces. Prompt and schema engineering is easy to deploy but often fragile under distribution shift and strict parsers, while continual parameter-efficient fine-tuning improves reliability at the cost of training, maintenance, and potential forgetting. We identify a critical Lazy Agent failure mode where tool necessity is nearly perfectly decodable from mid-layer activations, yet the model remains conservative in entering tool mode, revealing a representation-behavior gap. We propose Activation Steering Adapter (ASA), a training-free, inference-time controller that performs a single-shot mid-layer intervention and targets tool domains via a router-conditioned mixture of steering vectors with a probe-guided signed gate to amplify true intent while suppressing spurious triggers. On MTU-Bench with Qwen2.5-1.5B, ASA improves strict tool-use F1 from 0.18 to 0.50 while reducing the false positive rate from 0.15 to 0.05, using only about 20KB of portable assets and no weight updates.

cs.SE↗
arXiv2026

Hierarchical Adaptive Feature Refinement Network for VHR Remote Sensing Image Segmentation

Semantic segmentation of very-high-resolution (VHR) remote sensing imagery increasingly benefits from strong pretrained hierarchical encoders, yet exploiting their multi-stage representations remains difficult. Nearby regions demand different balances between fine detail and semantic context, aggressive task-specific transformations perturb useful pretrained features, and conventional semantic supervision provides limited structural guidance. We present HAFR-Net, a progressive refinement framework that adaptively organizes and conservatively refines hierarchical representations instead of replacing them with a monolithic decoder transformation. Heterogeneity-Guided Stage-Adaptive Fusion (HG-SAF) predicts dense stage weights conditioned on local feature variation. A Frequency-Residual Adapter (FRA) then injects frequency information through a bounded, zero-initialized residual branch that keeps the fused representation as its reference. A Confusion-Aware Tri-Prior Decoder (CATP) finally regularizes the prediction with boundary, objectness, and training-derived class-relation cues. Under a matched Swin-B training and single-scale inference protocol, HAFR-Net attains 84.12%, 87.86%, 55.17%, and 67.70% mIoU on ISPRS Vaihingen, ISPRS Potsdam, LoveDA, and OpenEarthMap, improving the matched UPerNet baseline by 0.55, 0.95, 1.55, and 1.84 percentage points, respectively. Controlled analyses further show consistent spatial reweighting beyond content-only routing, improved boundary and thin-structure accuracy over matched spatial and spectral alternatives, and reduced confusion on pre-declared class pairs.

cs.CV↗
arXiv2026

Contrastive Mask Fidelity: Reference-Free Auditing of Ground-Truth Masks in Remote Sensing Semantic Segmentation

Semantic segmentation models are trained and evaluated against human-drawn masks, yet remote-sensing annotations are often coarse, incomplete, or misaligned; high overlap scores may then reflect agreement with imperfect labels rather than faithfulness to the image, creating an evaluation paradox. We introduce Contrastive Mask Fidelity (CMF), a training-free, reference-free metric that scores competing class masks directly against image evidence. CMF composites keep and erase counterfactual views of each mask and asks a frozen vision-language judge whether class evidence is concentrated inside the mask and absent outside. We validate CMF on controlled mask corruptions, then audit 10,731 image-class pairs across ten remote-sensing benchmarks using candidate masks from Seg-Probe, a training-free open-vocabulary probe built on SegEarth-OV3 that outperforms prior baselines on nine of ten datasets. The audit reveals systematic, class-dependent annotation distortion: man-made classes such as buildings, roads, and cars favor the candidate mask on 62-85% of pairs, whereas ambiguous land cover more often favors human annotations. On a blinded three-annotator consensus, CMF matches expert judgment on 81% of pairs, exceeding keep-only scoring, model confidence, and a trained label-quality baseline. Finally, conservative class-wise arbitration yields supervision that improves cross-domain transfer over raw annotations and matched replacement controls, positioning CMF as a scalable tool for auditing ground truth rather than presuming it infallible.

cs.CV↗
arXiv2026

Quantum cluster algebra realization for stated ${\rm SL}_n$-skein algebras and rotation-invariant bases for polygons

We construct a quantum cluster structure on the skew-field of fractions ${\rm Frac}({\mathscr S}_ω(\mathfrak{S}))$ of the stated ${\rm SL}_n$-skein algebra ${\mathscr S}_ω(\mathfrak{S})$, where $\mathfrak{S}$ is a triangulable pb surface without interior punctures. This work complements the construction for the projected stated skein algebra $\widetilde{\mathscr S}_ω(\mathfrak{S})$ given by the last two authors. Let ${\mathscr S}_ω^{\rm fr}(\mathfrak{S})$ denote the localization of ${\mathscr S}_ω(\mathfrak{S})$ at the multiplicative set generated by all frozen variables. Let ${\mathscr A}_ω^{\rm fr}(\mathfrak{S})$ and ${\mathscr U}_ω^{\rm fr}(\mathfrak{S})$ (respectively $\overline{\mathscr A}_ω(\mathfrak{S})$ and $\overline{\mathscr U}_ω(\mathfrak{S})$) denote the quantum cluster algebra and quantum upper cluster algebra associated to ${\rm Frac}({\mathscr S}_ω(\mathfrak{S}))$ (respectively ${\rm Frac}(\widetilde{\mathscr S}_ω(\mathfrak{S}))$). We prove that \[ \widetilde{\mathscr S}_ω(\mathfrak{S}) = \overline{\mathscr A}_ω(\mathfrak{S}) = \overline{\mathscr U}_ω(\mathfrak{S}) \quad \text{and} \quad {\mathscr S}_ω^{\rm fr}(\mathfrak{S}) = {\mathscr A}_ω^{\rm fr}(\mathfrak{S}) = {\mathscr U}_ω^{\rm fr}(\mathfrak{S}) \] whenever $\mathfrak{S}$ is a polygon. As a consequence, when $\mathfrak{S}$ is a polygon, we show that the theta basis of $\overline{\mathscr U}_ω(\mathfrak{S})$ (respectively ${\mathscr U}_ω^{\rm fr}(\mathfrak{S})$) yields a rotation-invariant basis of $\overline{\mathscr S}_ω(\mathfrak{S})$ (respectively ${\mathscr S}_ω^{\rm fr}(\mathfrak{S})$) with several desirable properties, including positivity and a natural parametrization.

math.QA↗
arXiv2026

ST-Prune: Training-Free Spatio-Temporal Token Pruning for Vision-Language Models in Autonomous Driving

Vision-Language Models (VLMs) have become central to autonomous driving systems, yet their deployment is severely bottlenecked by the massive computational overhead of multi-view camera and multi-frame video input. Existing token pruning methods, primarily designed for single-image inputs, treat each frame or view in isolation and thus fail to exploit the inherent spatio-temporal redundancies in driving scenarios. To bridge this gap, we propose ST-Prune, a training-free, plug-and-play framework comprising two complementary modules: Motion-aware Temporal Pruning (MTP) and Ring-view Spatial Pruning (RSP). MTP addresses temporal redundancy by encoding motion volatility and temporal recency as soft constraints within the diversity selection objective, prioritizing dynamic trajectories and current-frame content over static historical background. RSP further resolves spatial redundancy by exploiting the ring-view camera geometry to penalize bilateral cross-view similarity, eliminating duplicate projections and residual background that temporal pruning alone cannot suppress. These two modules together constitute a complete spatio-temporal pruning process, preserving key scene information under strict compression. Validated across four benchmarks spanning perception, prediction, and planning, ST-Prune establishes new state-of-the-art for training-free token pruning. Notably, even at 90\% token reduction, ST-Prune achieves near-lossless performance with certain metrics surpassing the full-model baseline, while maintaining inference speeds comparable to existing pruning approaches.

cs.CV↗
arXiv2026

Orderings of Generalized k-Markov Numbers

A $k$-Markov number is a positive integer that appears in a positive integral solution to the Diophantine equation $x^2 + y^2 + z^2 + k(xy + xz + yz) = (3+3k)xyz$. This equation was introduced by Gyoda and Matsushita. When $k =0$, this definition recovers that of ordinary Markov numbers. The set of $k$-Markov numbers can be indexed by pairs of coprime positive integers. There is a consistent way to label non-coprime pairs with positive integers as well, yielding a larger set of ``generalized $k$-Markov numbers.'' In this paper, we classify lines along which the generalized $k$-Markov numbers grow monotonically, extending work in the ordinary case by Lee-Li-Rabideau-Schiffler and by the second author. We find that, as $k$ grows, the $k$-Markov numbers are more likely to be monotonic along a random line. This gives evidence that a $k$-version of Frobenius' uniqueness conjecture, which has been proposed by Gyoda and Maruyama, could be true.

math.NT↗
arXiv2026

CentaurTA Studio: A Self-Improving Human-Agent Collaboration System for Thematic Analysis

Thematic analysis is difficult to scale: manual workflows are labor-intensive, while fully automated pipelines often lack controllability and transparent evaluation. We present \textbf{CentaurTA Studio}, a web-based system for self-improving human--agent collaboration in open coding and theme construction. The system integrates (1) a two-stage human feedback pipeline separating simulator drafting and expert validation, (2) persistent prompt optimization that distills validated feedback into reusable alignment principles, and (3) rubric-based evaluation with early stopping for process control. Across three domains, CentaurTA achieves the strongest performance in both Open Coding and Theme Construction, reaching up to 92.12\% accuracy and consistently outperforming baseline systems. Agreement between the rubric-based LLM judge and human annotators reaches substantial reliability (average $κ= 0.68$). Ablation studies show that removing the feedback loop reduces performance from 90\% to 81\%, while eliminating the Critic or early stopping degrades accuracy or increases interaction cost. The full system reaches peak performance within 10 iterative rounds (about 25 minutes), demonstrating improved efficiency over expert-only refinement.

cs.HC↗
arXiv2026

DanceHA: A Multi-Agent Framework for Document-Level Aspect-Based Sentiment Analysis

Aspect-Based Sentiment Intensity Analysis (ABSIA) has garnered increasing attention, though research largely focuses on domain-specific, sentence-level settings. In contrast, document-level ABSIA--particularly in addressing complex tasks like extracting Aspect-Category-Opinion-Sentiment-Intensity (ACOSI) tuples--remains underexplored. In this work, we introduce DanceHA, a multi-agent framework designed for open-ended, document-level ABSIA with informal writing styles. DanceHA has two main components: Dance, which employs a divide-and-conquer strategy to decompose the long-context ABSIA task into smaller, manageable sub-tasks for collaboration among specialized agents; and HA, Human-AI collaboration for annotation. We release Inf-ABSIA, a multi-domain document-level ABSIA dataset featuring fine-grained and high-accuracy labels from DanceHA. Extensive experiments demonstrate the effectiveness of our agentic framework and show that the multi-agent knowledge in DanceHA can be effectively transferred into student models. Our results highlight the importance of the overlooked informal styles in ABSIA, as they often intensify opinions tied to specific aspects.

cs.CL↗
arXiv2026

ACE-Bench: A Lightweight Benchmark for Evaluating Azure SDK Usage Correctness

We present ACE-Bench (Azure SDK Coding Evaluation Benchmark), an execution-free benchmark that provides fast, reproducible pass or fail signals for whether large language model (LLM)-based coding agents use Azure SDKs correctly-without provisioning cloud resources or maintaining fragile end-to-end test environments. ACE-Bench turns official Azure SDK documentation examples into self-contained coding tasks and validates solutions with task-specific atomic criteria: deterministic regex checks that enforce required API usage patterns and reference-based LLM-judge checks that capture semantic workflow constraints. This design makes SDK-centric evaluation practical in day-to-day development and CI: it reduces evaluation cost, improves repeatability, and scales to new SDKs and languages as documentation evolves. Using a lightweight coding agent, we benchmark multiple state-of-the-art LLMs and quantify the benefit of retrieval in an MCP-enabled augmented setting, showing consistent gains from documentation access while highlighting substantial cross-model differences.

cs.DC↗
arXiv2026

A mutation invariant for skew-symmetrizable matrices

Matrix mutation of skew-symmetrizable matrices is foundational in cluster algebra theory. Effective mutation invariants are essential for determining whether two matrices lie in the same mutation class. Casals~\cite{Casals} introduced a binary mutation invariant for skew-symmetric matrices. In this paper, we extend Casals' construction to the skew-symmetrizable setting. When the skew-symmetrizer $d_1,\dots, d_n$ is pairwise coprime, we obtain two distinct extensions of this invariant.

math.CO↗
arXiv2026

Regulatory Hub Discovery in MDD Methylome: Hypotheses for Molecular Subtypes via Computational Analysis

Major Depressive Disorder (MDD) is a clinically heterogeneous syndrome with diverse etiological pathways. Traditional Epigenome-Wide Association Studies (EWAS) have successfully identified risk loci based on differential methylation magnitude. As a complementary perspective, effect-size-based ranking alone may not fully capture regulatory nodes that exhibit modest methylation changes but occupy critical upstream positions in biological networks. Here, we report findings and hypotheses from a two-tier computational analysis of DNA methylation data (GSE198904; \(n=206\) ), combining conventional statistical approaches with machine learning-assisted regulatory inference.

cs.CE↗
arXiv2025

SpikGPT: A High-Accuracy and Interpretable Spiking Attention Framework for Single-Cell Annotation

Accurate and scalable cell type annotation remains a challenge in single-cell transcriptomics, especially when datasets exhibit strong batch effects or contain previously unseen cell populations. Here we introduce SpikGPT, a hybrid deep learning framework that integrates scGPT-derived cell embeddings with a spiking Transformer architecture to achieve efficient and robust annotation. scGPT provides biologically informed dense representations of each cell, which are further processed by a multi-head Spiking Self-Attention mechanism for energy-efficient feature extraction. Across multiple benchmark datasets, SpikGPT consistently matches or exceeds the performance of leading annotation tools. Notably, SpikGPT uniquely identifies unseen cell types by assigning low-confidence predictions to an "Unknown" category, allowing accurate rejection of cell states absent from the training reference. Together, these results demonstrate that SpikGPT is a versatile and reliable annotation tool capable of generalizing across datasets, resolving complex cellular heterogeneity, and facilitating discovery of novel or disease-associated cell populations.

q-bio.QM↗
arXiv2025

The First Scientific Flight and Observations of the 50-mm Balloon-Borne White-Light Coronagraph

A 50-mm balloon-borne white-light coronagraph (BBWLC) to observe whitelight solar corona over the altitude range from 1.08 to 1.50 solar radii has recently been indigenously developed by Yunnan Observatories in collaboration with Shangdong University (in Weihai) and Changchun Institute of Optics, Fine Mechanics and Physics, which will significantly improve the ability of China to detect and measure inner corona. On 2022 October 4, its first scientific flight took place at the Dachaidan area in Qinghai province of China. We describe briefly the BBWLC mission including its optical design, mechanical structure, pointing system, the first flight and results associated with the data processing approach. Preliminary analysis of the data shows that BBWLC imaged the Kcorona with three streamer structures on the west limb of the Sun. To further confirm the coronal signals obtained by BBWLC, comparisonswere made with observations of the Kcoronagraph of the High Altitude Observatory and the Atmospheric ImagingAssembly on board the Solar Dynamics Observatory. We conclude that BBWLC eventually observed the white-light corona in its first scientific flight.

astro-ph.SR↗