arXiv subjects

Junjie Huang

Publications and source records attributed to Junjie Huang.

At least 19 recordsLinked to original sources

arXiv2026

Turbulence Cascade in Cygnus X Revealed by Multi-point VDF Method

Turbulence plays a crucial role in regulating star formation activities within molecular clouds, yet few methods can directly reveal its properties and underlying processes. We use molecular line data from the Nobeyama 45m Cygnus X CO Survey to study the turbulence properties and their relationships with star-forming activities and/or other non-thermal motions. In this work, we apply the multi-point velocity dispersion function (VDF), rather than direct linewidth measurements, to investigate the non-thermal properties of molecular cloud motions. We filter out the large-scale ordered structure and isolate a relatively small-scale turbulence component. Through the Friends In Velocity (FIVe) algorithm, we identify 10 substructures of the clouds and derive the turbulent properties of each cloud using the VDF method. We find that both the cloud-complex regions and the 10 velocity substructures exhibit turbulence correlation lengths of $\sim 2$--5 pc. This plateau scale suggests a parsec-scale turbulence correlation or driving scale in Cygnus X. Below this scale, the rising VDFs trace the velocity scaling of the turbulent cascade, whereas larger-scale VDF variations likely reflect cloud-scale motions. The comparison between cloud complexes and substructures further suggests that, in observational data, the VDF may constrain the turbulence correlation scale more robustly than the turbulence velocity dispersion.

astro-ph.GA↗

arXiv2026

From Final Artifacts to Trajectories: Retrospective Process Supervision for Evidence-Grounded Long-Form Generation

Trajectory data is getting more vital for training large language models for boosting the agentic abilities. Unlike the verifiable domains such as coding or mathematics, scaling trajectory data for open-ended tasks is much more difficult because these tasks lack singular ground truth and are costly to annotate or verify. In this paper, we propose RetroGen, a self-improving framework of retrospective process supervision. Our key observation is that although expert trajectories are scarce, high-quality final artifacts such as literature reviews, analyst reports and legal judgments, are abundant in pre-training data and can be viewed as compressed traces of the evidence-seeking processes that produced them. RetroGen reconstructs candidate latent trajectories from expert artifacts, verifies them against both the artifact and supporting evidence, and trains models on their own successful reconstruction data, without requiring trajectory data from stronger models. Experiments show that RetroGen improves grounding, faithful synthesis, and long-form evidence-seeking agent tasks.

cs.CL↗

arXiv2026

SLAI T-Rex: Full-Parameter Post-training of the DeepSeek-V4 Family on Ascend SuperPOD

Full-parameter post-training of trillion-parameter-scale MoE models introduces substantial system-level challenges for large-scale distributed training, including severe memory pressure, non-overlapped communication overhead, and inefficient kernel execution. While most large-scale LLM training systems are built around GPU-based clusters, this report presents an end-to-end optimization practice on the Ascend NPU SuperPOD. Using the DeepSeek-V4 model family as the target workload, we develop a hierarchical optimization framework spanning model-level parallelism, computation-communication orchestration, and low-level kernel execution. The resulting system achieves 34.22% Model FLOPs Utilization (MFU) with a 2.93x improvement over the open-source baseline recipe while maintaining training stability. Building on this optimized infrastructure, we further establish a CPT and SFT workflow for complex Operations Research (OR) tasks. We refer to the integrated framework as SLAI T-Rex. Using DeepSeek-V4-Flash, we develop OR-oriented CPT and SFT data pipelines that combine collected domain resources with solver-verified synthetic optimization documents. The resulting dataset contains 10K high-quality SFT samples spanning four task categories and three problem representations. The specialized model achieves the highest average zero-shot Pass@1 score among the evaluated models, reaching 71.81% and outperforming GPT-5.4-Mini and the base DeepSeek-V4-Flash model by 3.98 and 11.27 percentage points, respectively. Overall, this work demonstrates a full-stack pathway from efficient trillion-parameter model post-training on Ascend infra to domain-specialized Flash models for solver-grounded mathematical modeling, advancing frontier-model systems for complex reasoning.

cs.CL↗

arXiv2026

ForestBench: A Unified Graph Framework for Evaluating Multi-Agent Collaboration

Multi-agent systems (MAS) built on Large Language Models (LLMs) are proliferating rapidly, but their heterogeneous execution traces provide no common basis for evaluation across methods. Outcome-only benchmarks discard collaborations, whereas LLM-as-Judge evaluation requires additional, model-dependent inference and can vary with the LLM and rubric. We introduce a generalizable evaluation framework that maps native MAS traces into a shared space of unified collaboration graphs, enabling different methods to be evaluated under the same representation, reference set, and metric panel. Candidate graphs are compared with a query-specific reference forest. Each forest is a benchmark-provided collection of verified-success graphs: it records diverse ways in which representative MAS methods can complete the task, rather than prescribing a unique optimal process. Instantiating the framework as ForestBench, we filter $844$ collaboration-necessary queries from seven public datasets, precompute ten successful target-conditioned reference graphs per query, and evaluate six representative MAS frameworks. Controlled backbone, reference-construction, and perturbation studies test the stability and scope of evaluation. Once the benchmark forests are built, ForestBench scores a trace in milliseconds without further LLM inference, providing a reusable structural basis for comparing diverse MAS collaboration traces.

cs.AI↗

arXiv2026

NGM-RAG: Neural Graph Matching based Retrieval-Augmented Generation

Retrieval-Augmented Generation (RAG) significantly enhances the ability of Large Language Models (LLMs) to provide accurate and contextually relevant answers by dynamically integrating external databases. However, traditional RAG methods are primarily constrained by their reliance on text-based retrieval strategies, which often struggle with complex questions requiring multi-hop reasoning. To address this limitation, we introduce Neural Graph Matching based Retrieval-Augmented Generation (NGM-RAG), a novel framework that leverages graph structures to effectively capture and utilize relational knowledge for improved retrieval and answer generation. NGM-RAG explicitly incorporates graph construction, graph matching, and answer generation into a unified process. Within this framework, we propose a neural graph matching approach that combines text-based matching with Graph Neural Networks (GNNs). By employing an adaptive weighting strategy, NGM-RAG efficiently integrates multiple matching methods to select the most relevant contextual node information for answer generation. Experimental results on multi-hop question answering and long-context summarization tasks demonstrate that our NGM-RAG model achieves superior performance compared to both traditional NaiveRAG methods and state-of-the-art graph-enhanced approaches such as GraphRAG and LightRAG.

cs.IR↗

arXiv2026

Post-starburst Galaxies with Active Galactic Nucleus: Properties and Evolutionary Sequences

Post-starburst (PSB) galaxies, identified by strong Balmer absorption and weak nebular emission, provide a key laboratory for studying rapid quenching. Using the final data release of the SDSS-IV MaNGA survey, we follow the traditional PSB selection criteria of Chen et al. (2019) and develop a new method to identify regions that simultaneously exhibit PSB features and nuclear activities (AGN-PSBs). Our final sample comprises 48 AGN-PSBs, 92 central PSBs (CPSBs), 89 ring-like PSBs (RPSBs), and 828 irregular PSBs (IPSBs). We find the global and spatially resolved properties of CPSBs and RPSBs are consistent with the results of Chen et al. (2019). In this work, we focus on the properties of AGN-PSBs, comparing them with CPSBs, RPSBs, and control galaxies. Similar to CPSBs and RPSBs, AGN-PSBs show positive $\mathrm{D}_{n}4000$ gradients relative to negative $\mathrm{D}_{n}4000$ gradients of their controls, which indicates younger stellar populations in the central region than that in the outskirt. Among the three sub-types, high-mass CPSBs (H-CPSBs, with $\log(M_{*}/M_{\odot})>9.5$) display the highest incidence of merger remnants and gas--star kinematic misalignment, consistent with a merger/interaction-dominated origin. AGN-PSBs and RPSBs, however, show lower and comparable fractions of merger remnants and gas--star kinematic misalignment, favoring less violent external mechanisms. Based on radial profiles of mass-weighted age and $V_{\rm star}/σ_{\rm star}$, we suggest that RPSBs can evolve into AGN-PSBs, whereas H-CPSBs likely follow a distinct evolutionary pathway. The existence of RPSBs and IPSBs also indicates that AGN feedback is not a necessary condition for the formation of PSB.

astro-ph.GA↗

arXiv2026

MicroAgent: Context-Augmented Multi-Agent Framework for Automatic Microservice Decomposition

The adoption of Microservice Architecture (MSA) has revolutionized software engineering by enhancing scalability, agility, and maintainability over traditional monolithic applications. As more developers transition their legacy systems to microservice-based architectures, effective microservice decomposition-partitioning monolithic applications into highly cohesive services-becomes vital. However, this decomposition task presents significant challenges. Manual approaches are time-consuming and labor-intensive. Existing automated methods often fail to capture the necessary semantic insights from complex applications, while naive applications of Large Language Models tend to overlook crucial contextual information and design principles, leading to suboptimal results. To address these challenges, we propose MicroAgent, a Context-Augmented Multi-Agent Framework for Microservice Decomposition. Our framework divides the decomposition process into five distinct subtasks and assigns each to a specialized agent. To enhance the effectiveness of each agent, we provide tailored, multi-granularity context that keeps its analysis focused and mitigates information overload. Furthermore, to ensure the decomposition adheres to established design principles, we integrate analytical tools that guide the agents' decision-making. Experimental evaluations on 10 Java Web applications demonstrate that MicroAgent achieves an average decomposition accuracy of 89.2%, outperforming the state-of-the-art method by 24.6%. We also conduct a case study to highlight the practical benefits of our design.

cs.SE↗

arXiv2026

On the Maximal Length of MDS Elliptic Codes

The determination of the maximal length of maximum distance separable (MDS) codes arising from elliptic curves is a central problem in coding theory. For an elliptic curve $E$ over $\mathbb{F}_q$, let $\operatorname{MEC}(k,q)$ denote the maximal length of a $q$-ary MDS elliptic code of dimension $k$. It was recently shown that $\operatorname{MEC}(k,q)\le\frac{q+1}{2}+\sqrt{q}$ for $q\ge289$ and $3\le k\le(q+1-2\sqrt{q})/10$, with equality for odd $k$ when $q$ is an odd square. This paper investigates the remaining open cases, namely even dimension $k$, non-square $q$ and fields of characteristic $2$, and provides a complete resolution of the tightness question for the two natural parity regimes of $q+1+\lfloor 2\sqrt{q}\rfloor$. We prove that if the support of $G$ (used to define the code) consists of $\mathbb{F}_q$-rational points, the bound decreases to $\frac{q+1}{2}+\sqrt{q}-1$ for even $k$. Without this restriction, we construct MDS codes attaining $\frac{q+1}{2}+\sqrt{q}$ for even $k$. More generally, we establish $\operatorname{MEC}(k,q)=\frac{q+1+\lfloor2\sqrt{q}\rfloor}{2}$ when $q+1+\lfloor2\sqrt{q}\rfloor$ is even, and $\operatorname{MEC}(k,q)=\frac{q+\lfloor2\sqrt{q}\rfloor}{2}$ when it is odd.

cs.IT↗

arXiv2026

On Permutation Groups of Cyclic Codes over Finite Fields

The permutation groups of cyclic codes are widely applicable in determining the weight distribution of codes, decoding theory and various other areas. In this paper, by employing two distinct matrix representations, we can relate cyclic codes with very long lengths and special generator polynomials to those with prime lengths. Consequently, we mainly determine the permutation groups of certain cyclic codes over $\mathbb{F}_{r^α}$ with lengths $hp$, $r^mp^n$ and $pq$ and special generator polynomials where $h$ is a positive integer and $p$, $q$ and $r$ are distinct prime numbers. For length $pq$, we manage to provide the permutation groups of cyclic codes with generator polynomials $Q_{pq}(x)$(the $pq$-th cyclotomic polynomial) or others, which seems to be the first work about permutation groups of cyclic codes with generator polynomials that are factors of $x^{pq}-1$ but not factors of $x^p-1(\text{or }x^q-1)$.

cs.IT↗

arXiv2026

Locally Repairable Codes with Availability via Elliptic Function Fields

Locally repairable codes with availability have become essential components in modern large-scale distributed cloud storage systems and numerous other applications. In this paper, we focus on the construction of locally repairable codes with one or two recovering sets via elliptic function fields. Prior pioneering work by Li et al. (IEEE Trans. Inf. Theory, vol. 65, no. 1, 2019) and Ma and Xing (J. Comb. Theory Ser. A., vol. 193, 2023) employed maximal supersingular elliptic curves to obtain several optimal (classical) locally repairable codes. In contrast, we consider ordinary elliptic curves with many rational points. This approach yields several new families of $q$-ary optimal locally repairable codes with length $O(q+2\sqrt{q})$ and flexible locality. Consequently, our work broadens the selection of curves available for the construction of optimal locally repairable codes. Furthermore, we present a general framework for constructing locally repairable codes with two recovering sets via automorphism groups of elliptic function fields. To realize this framework, we devise a novel construction for determining the functions $e_i$ in the construction of locally repairable codes. By employing both supersingular and ordinary elliptic curves, we obtain several families of locally repairable codes with two recovering sets. In particular, we construct a family of $q^2$-ary locally repairable codes with two recovering sets, achieving length $O(q^2+2q)$ and Singleton-defect $O\!\left(\frac{2\ell}{q^2+2q-8\ell}\right)$, where $\ell \mid\mid q + 2$ with $4\ell < q$.

cs.IT↗

arXiv2026

RABC-Net: Reliability-Aware Annotation-Free Skin Lesion Segmentation for Low-Resource Dermoscopy

Pixel-level annotation is costly in low-resource dermoscopy. We present RABC-Net, a reliability-aware annotation-free segmentation system that combines pseudo-label reliability learning, restricted target-domain adaptation, and Reliability-Adaptive Boundary Calibration (RABC). The system decouples reliability learning from deployment: uncertainty-aware pseudo-label interaction shapes robust representations during training, while the image-only inference path is preserved and RABC performs local logit-space calibration from boundary confidence, uncertainty, and foreground probability. No manual masks are used for training or target-domain adaptation; validation labels, when available, are used only for final operating-point selection. Across ISIC-2017, ISIC-2018, and PH2, RABC-Net achieves macro-average DICE/JAC of 86.58\%/79.47\% and consistent matched-protocol results. Controlled within-study analyses show that RABC provides localized gains over nonlearned boundary correction, while the overall result comes from the full reliability-aware system. Adaptation updates only 3.50\% of model parameters, image-only inference runs at 87.4 FPS, and the selected operating points use $σ=0$ on all three datasets, indicating that learned calibration avoids extra smoothing at deployment.

cs.CV↗

arXiv2026

Inductive Dual-Polarity Modeling via Static-Dynamic Disentanglement for Dynamic Signed Networks

Dynamic signed networks (DSNs) are common in online platforms, where time-stamped positive and negative relations evolve over time. A core task in DSNs is dynamic edge prediction, which forecasts future relations by jointly modeling edge existence and polarity (positive, negative, or non-existent). However, existing dynamic signed network embedding (DSNE) methods often entangle positive and negative signals within a shared temporal state and rely on node-specific temporal trajectories, which can obscure polarity-asymmetric dynamics and harm inductive generalization, especially under cold-start evaluation. We study an inductive setting where each test edge contains at least one endpoint node held out from training, while its interactions prior to the prediction time are available as historical evidence. The model must therefore infer representations for unseen nodes solely from such limited history. We propose IDP-DSN, an Inductive Dual-Polarity framework for Dynamic Signed Networks. IDP-DSN maintains sign-selective memories to model positive and negative temporal dynamics separately, performs history-only neighborhood inference for unseen nodes (instead of learned node-wise trajectories), and enforces polarity-wise static--dynamic disentanglement via an orthogonality regularizer. Experiments on BitcoinAlpha, BitcoinOTC, Wiki-RfA, and Epinions demonstrate consistent improvements over the strongest baselines, achieving relative Macro-F1 gains of 16.8/23.4%, 16.9/24%, 30.1/25.5%, and 18.7/28.9% in the transductive/inductive settings, respectively. These results highlight the effectiveness of IDP-DSN on DSNs, particularly under inductive cold-start evaluation for dynamic signed edge prediction.

cs.SI↗

arXiv2026

Agent-World: Scaling Real-World Environment Synthesis for Evolving General Agent Intelligence

Large language models are increasingly expected to serve as general-purpose agents that interact with external, stateful tool environments. The Model Context Protocol (MCP) and broader agent skills offer a unified interface for connecting agents with scalable real-world services, but training robust agents remains limited by the lack of realistic environments and principled mechanisms for life-long learning. In this paper, we present \textbf{Agent-World}, a self-evolving training arena for advancing general agent intelligence through scalable environments. Agent-World has two main components: (1) Agentic Environment-Task Discovery, which autonomously explores topic-aligned databases and executable tool ecosystems from thousands of real-world environment themes and synthesizes verifiable tasks with controllable difficulty; and (2) Continuous Self-Evolving Agent Training, which combines multi-environment reinforcement learning with a self-evolving agent arena that automatically identifies capability gaps through dynamic task synthesis and drives targeted learning, enabling the co-evolution of agent policies and environments. Across 23 challenging agent benchmarks, Agent-World-8B and 14B consistently outperforms strong proprietary models and environment scaling baselines. Further analyses reveal scaling trends in relation to environment diversity and self-evolution rounds, offering insights for building general agent intelligence.

cs.AI↗

arXiv2026

Can LLMs Help Decentralized Dispute Arbitration? A Case Study of UMA-Resolved Markets on Polymarket

Web3 prediction markets, exemplified by Polymarket, have gained prominence for leveraging collective intelligence to forecast a wide range of social, political, and sports events. However, among the thousands of prediction market events, consensus disputes still arise due to imperfections in market mechanisms. On Polymarket alone, the trading volume involving disputed events has reached $972,370,804.71, underscoring the critical need for objective and efficient dispute resolution. In this study, we introduce large language models (LLMs) to: (1) evaluate whether web-enabled LLMs can reproduce the decision quality of UMA's on-chain voting process once a dispute has been raised, and (2) predict, based on event rules, which market events are likely to face future disputes before they occur. Our findings show that LLMs are unable to reliably predict which events will become disputed in advance; however, once a dispute is initiated, web-enabled LLMs achieve 89.58% agreement with UMA's final resolutions and demonstrate strong stability.

cs.CY↗

arXiv2026

Embodied Science: Closing the Discovery Loop with Agentic Embodied AI

Artificial intelligence has demonstrated remarkable capability in predicting scientific properties, yet scientific discovery remains an inherently physical, long-horizon pursuit governed by experimental cycles. Most current computational approaches are misaligned with this reality, framing discovery as isolated, task-specific predictions rather than continuous interaction with the physical world. Here, we argue for embodied science, a paradigm that reframes scientific discovery as a closed loop tightly coupling agentic reasoning with physical execution. We propose a unified Perception-Language-Action-Discovery (PLAD) framework, wherein embodied agents perceive experimental environments, reason over scientific knowledge, execute physical interventions, and internalize outcomes to drive subsequent exploration. By grounding computational reasoning in robust physical feedback, this approach bridges the gap between digital prediction and empirical validation, offering a roadmap for autonomous discovery systems in the life and chemical sciences.

cs.AI↗

arXiv2026

Accelerating Benchmarking of Functional Connectivity Modeling via Structure-aware Core-set Selection

Benchmarking the hundreds of functional connectivity (FC) modeling methods on large-scale fMRI datasets is critical for reproducible neuroscience. However, the combinatorial explosion of model-data pairings makes exhaustive evaluation computationally prohibitive, preventing such assessments from becoming a routine pre-analysis step. To break this bottleneck, we reframe the challenge of FC benchmarking by selecting a small, representative core-set whose sole purpose is to preserve the relative performance ranking of FC operators. We formalize this as a ranking-preserving subset selection problem and propose Structure-aware Contrastive Learning for Core-set Selection (SCLCS), a self-supervised framework to select these core-sets. SCLCS first uses an adaptive Transformer to learn each sample's unique FC structure. It then introduces a novel Structural Perturbation Score (SPS) to quantify the stability of these learned structures during training, identifying samples that represent foundational connectivity archetypes. Finally, while SCLCS identifies stable samples via a top-k ranking, we further introduce a density-balanced sampling strategy as a necessary correction to promote diversity, ensuring the final core-set is both structurally robust and distributionally representative. On the large-scale REST-meta-MDD dataset, SCLCS preserves the ground-truth model ranking with just 10% of the data, outperforming state-of-the-art (SOTA) core-set selection methods by up to 23.2% in ranking consistency (nDCG@k). To our knowledge, this is the first work to formalize core-set selection for FC operator benchmarking, thereby making large-scale operators comparisons a feasible and integral part of computational neuroscience. Code is publicly available on https://github.com/lzhan94swu/SCLCS

cs.LG↗

arXiv2026

ReMiT: RL-Guided Mid-Training for Iterative LLM Evolution

Standard training pipelines for large language models (LLMs) are typically unidirectional, progressing from pre-training to post-training. However, the potential for a bidirectional process--where insights from post-training retroactively improve the pre-trained foundation--remains unexplored. We aim to establish a self-reinforcing flywheel: a cycle in which reinforcement learning (RL)-tuned model strengthens the base model, which in turn enhances subsequent post-training performance, requiring no specially trained teacher or reference model. To realize this, we analyze training dynamics and identify the mid-training (annealing) phase as a critical turning point for model capabilities. This phase typically occurs at the end of pre-training, utilizing high-quality corpora under a rapidly decaying learning rate. Building upon this insight, we introduce ReMiT (Reinforcement Learning-Guided Mid-Training). Specifically, ReMiT leverages the reasoning priors of RL-tuned models to dynamically reweight tokens during the mid-training phase, prioritizing those pivotal for reasoning. Empirically, ReMiT achieves an average improvement of 3\% on 10 pre-training benchmarks, spanning math, code, and general reasoning, and sustains these gains by over 2\% throughout the post-training pipeline. These results validate an iterative feedback loop, enabling continuous and self-reinforcing evolution of LLMs.

cs.CL↗

arXiv2026

Successive Cancellation List Decoding of Extended Reed-Solomon Codes

Reed-Solomon (RS) codes are an important class of non-binary error-correction codes. They are particularly competent in correcting burst errors, being widely applied in modern communications and data storage systems. This also thanks to their distance property of reaching the Singleton bound, being the maximum distance separable (MDS) codes. This paper proposes a new list decoding for extended RS (eRS) codes defined over a finite field of characteristic two, i.e., F_{2^n}. It is developed based on transforming an eRS code into n binary polar codes. Consequently, it can be decoded by the successive cancellation (SC) decoding and further their list decoding, i.e., the SCL decoding. A pre-transformed matrix is required for reinterpretating the eRS codes, which also determines their SC and SCL decoding performances. Its column linear independence property is studied, leading to theoretical characterization of their SC decoding performance. Our proposed decoding and analysis are validated numerically.

cs.IT↗