Searcharxiv⌕ Search

arXiv subjects

Bo Huang

Publications and source records attributed to Bo Huang.

At least 19 recordsLinked to original sources
arXiv2026

Long-window 4DVar for reanalysis using a differentiable weather model

Atmospheric reanalyses combine observations with model forecasts using complex data assimilation systems. We test whether a differentiable weather model permits a simpler and more accurate method based on a long-window four-dimensional variational data assimilation (4D-Var) formulation that omits the conventional background-error term. The method uses automatic differentiation to find optimal NeuralGCM initial conditions that minimize the misfit to real surface-pressure observations distributed across overlapping windows of two to seven days, assuming no model error. Cycling at 6-hour intervals for three months beginning 1 January 2015 yields a stable reanalysis with smaller error relative to ERA5 in 500-hPa geopotential height than the Twentieth Century Reanalysis version 3 (20CRv3), which uses an ensemble Kalman filter to assimilate the same observations. Every window produces smaller errors than 20CRv3, with analysis error for the four-day window approximately 55% smaller than for 20CRv3. At the end of the four-day window, which does not benefit from future observations, error remains approximately 38% smaller than 20CRv3. Analyses degrade slightly beyond four days, which we attribute to the increasing importance of model error.

physics.ao-ph↗
arXiv2026

KAT-Coder-V2.5 Technical Report

We present KAT-Coder-V2.5, a coding-focused agentic model trained to act autonomously inside real, executable repositories rather than as a single-turn code generator. Its capability is bottlenecked less by model scale than by the scarcity of reproducible environments, verifiable rewards, and high-value trajectories, which we address with an end-to-end agentic post-training framework. AutoBuilder reconstructs multilingual repositories into sandboxed environments with fail-to-pass and pass-to-pass verification at scale, from which we regenerate self-contained task specifications, recover near-miss trajectories, and distill supervision through process-aware filtering, while KwaiClawEnv synthesizes large-scale tool-use trajectories from executable services and real task seeds. We further scale reinforcement learning with harness randomization, a reliability-hardened sandbox, an asymmetric actor--critic PPO with hindsight-augmented value estimation, and a harness-oriented reward framework, and unify SWE, Agent-Claw, and WebCoding experts via Multi-Teacher On-Policy Distillation. Across six software-engineering and agentic benchmarks, KAT-Coder-V2.5 delivers the best agentic tool-use result on PinchBench and ranks second only to the frontier Opus 4.8 on repository-level software engineering. Our service is available at https://streamlake.com/product/kat-coder.

cs.SE↗
arXiv2026

Clarus: Coordinating Autonomous Research Agents toward Web-Scale Scientific Collaboration

Existing autonomous research agents can support parts of the research process, but most systems still treat research as either an isolated assistant task or a closed workflow. Therefore, autonomous science needs a collaboration infrastructure that coordinates projects, agents, and digital and physical resources. We identify this as a shift from code-centered execution loops to research-oriented collaboration processes, where questions, evidence, participants, and resources must be coordinated under uncertainty. In this framing, an agent may be an AI system, a human researcher, a team, a laboratory, or an organization-backed participant. To this end, we present Clarus, a collaboration infrastructure for coordinating autonomous research agents toward web-scale scientific collaboration. Clarus reformulates research as an open, auditable, attributable, and resource-aware multi-phase collaboration process. It defines a minimal project-agent-resource object model and organizes scientific collaboration through four layers including Research Application, Digital Collaboration, Physical Substrate, and Physical World. Core modules are implemented as pluggable mechanisms, allowing Clarus to adapt to task risk, collaboration structure, and resource constraints. Through a controlled paper-generation case study, we show that Clarus can organize a research goal into a traceable, reviewable, attributable, and accumulative collaboration network across phases, tasks, and participants. Together, the object model, collaboration protocol, trust mechanisms, and prototype validation provide an initial foundation for open research networks. Clarus is now available at clarus.holosai.io.

cs.AI↗
arXiv2026

SkillJuror: Measuring How Agent Skill Organization Changes Runtime Behavior

Agent Skills augment large language model (LLM) agents with procedural knowledge at inference time, but current benchmarks rarely distinguish what a Skill says from how it is organized. We study this distinction through Progressive Disclosure, where a concise root file points agents to supporting resources on demand, and compare it with a normalized flat baseline. We present SkillJuror, a framework for evaluating Skill writing paradigms through semantically controlled variants, matched multi-trial evaluations, and trajectory evidence while holding task knowledge fixed. In an 82-task SkillsBench study, Progressive Disclosure changes runtime behavior before aggregate outcomes: distinct Skill resources touched per trajectory rise from 1.18 to 3.85, and effective uptake events rise from 1.33 to 3.92. It also yields 17 additional verifier-passing trials out of 410 matched trials (+4.1%) over the normalized flat baseline. The benefit is task-dependent. Progressive Disclosure helps when supporting resources guide implementation, checking, or repair, but is weaker when success hinges on exact output conventions, numerical thresholds, or long artifact-generation pipelines. These results show that Skill organization is not mere presentation: it can change how agents search and apply procedural knowledge, while outcome gains depend on whether the exposed resources are actionable for the task. Code is available at https://github.com/zhiyuchen-ai/skill-juror.

cs.AI↗
arXiv2026

Magneto-Gravitational Regulated Streamer Accretion onto a Class 0 Protostellar System

How do magnetic fields shape the way young stars gather gas from their birth clouds? Using high-resolution Atacama Large Millimeter/submillimeter Array observations of a young triple protostellar system HOPS-182, we identify an elongated stream of gas, or accretion streamer, that extends over several thousand astronomical units (1 astronomical unit is the Earth-Sun distance) and carries a substantial flow of material toward the system. The gas speeds along this filament increase toward the star in a way consistent with gravitational free-fall, while the streamer's shape closely follows the magnetic field threading the region. By comparing the strengths of gravity and magnetic tension and measuring how the gas rotates compared with the local magnetic field, we show that the field is strong enough to help confine and guide the infalling gas and efficiently remove angular momentum. These results suggest that a substantial fraction of the material falling onto young protostellar systems can be funneled through elongated, magnetically structured accretion streamers.

astro-ph.GA↗
arXiv2026

From Holo Pockets to Electron Density: GPT-style Drug Design with Density

Recent advances in generative modeling have enabled significant progress in structure-based drug design (SBDD). Existing methods typically condition molecule generation on empty binding pockets from holo complexes, overlooking informative components such as the filler (ligands and solvent). Here, we leverage low-resolution electron density (ED) derived from the filler as a physically grounded condition for \textit{de novo} drug design. We consider two types of ED, calculated and cryo-EM/X-ray, obtainable from computational or experimental sources, supporting unified pre-training and experimental integration. Compared with rigid pocket representations, experimental ED naturally captures conformational flexibility and provides a more faithful description of the binding environment. Based on this, we introduce EDMolGPT, a decoder-only autoregressive framework that generates molecules from low-resolution ED point clouds. By grounding generation in physically meaningful density signals, EDMolGPT mitigates structural bias and produces molecules with 3D conformations. Evaluations on 101 biological targets verify the effectiveness. Our project page: https://jiahaochen1.github.io/EDMolGPT_Page/.

cs.AI↗
arXiv2026

Dynamic Prediction of the Target Survival Time in Metastatic Solid Tumor Cancer Clinical Trials

Overall survival (OS) is the gold standard for assessing patient benefit and cost-effectiveness of new cancer drugs. However, it is often difficult to use OS as the primary endpoint in randomized clinical trials (RCTs) for patients with metastatic cancer due to multiple reasons. In recent years, progression-free survival (PFS) has increasingly been used as the primary endpoint in metastatic cancer RCTs to accelerate development. However, regulatory authorities often seek mature OS data for approval. Therefore, it is critical to determine the target time when OS data are expected to be mature for reliable statistical inference. Motivated by an advanced renal cell carcinoma (RCC) clinical trial, we develop and investigate different prediction models leveraging information from disease progression to improve target OS prediction times. We propose a multivariate joint modeling approach considering components of progression and OS and extend three models commonly used for association to be used for OS prediction. To the best of our knowledge, this is the first comprehensive statistical study exploring the prediction of OS using different levels of information on disease progression and illustrating these models using a real, complex dataset. Our findings have significant implications for OS prediction.

stat.ME↗
arXiv2026

IDEA: An Interpretable and Editable Decision-Making Framework for LLMs via Verbal-to-Numeric Calibration

Large Language Models are increasingly deployed for decision-making, yet their adoption in high-stakes domains remains limited by miscalibrated probabilities, unfaithful explanations, and inability to incorporate expert knowledge precisely. We propose IDEA, a framework that extracts LLM decision knowledge into an interpretable parametric model over semantically meaningful factors. Through joint learning of verbal-to-numerical mappings and decision parameters via EM, correlated sampling that preserves factor dependencies, and direct parameter editing with mathematical guarantees, IDEA produces calibrated probabilities while enabling quantitative human-AI collaboration. Experiments across five datasets show IDEA with Qwen-3-32B (78.6%) outperforms DeepSeek R1 (68.1%) and GPT-5.2 (77.9%), achieving perfect factor exclusion and exact calibration -- precision unattainable through prompting alone. The implementation is publicly available at https://github.com/leonbig/IDEA.

cs.AI↗
arXiv2026

SW-$A^2$-Bench: Benchmarking Autonomous Software Agent Generation for Agentic Web

The Agentic Web is emerging as a paradigm in which autonomous software agents interact with online resources and with each other to accomplish user goals. However, the capacity of Agentic Web is still limited by insufficient autonomous software agent population, which has become a crucial challenge for scaling Agentic Web. In order to alleviate this, we study the task of automatically converting existing code repositories into autonomous software agents via coding agents, decompose the process into critical stages, and identify key technical hurdles. To systematically evaluate this capability, we propose SoftWare Agent generation for Agentic Web Bench (SW-$A^2$-Bench), the first benchmark designed for software agent generation. SW-$A^2$-Bench evaluates not only whether software agents can be generated, but also whether generated software agents are faithful to the source repositories and interoperable with other agents in multi-agent workflows. Our experiments demonstrate that our approach effectively activates the functional capabilities of code repositories and enables interoperable multi-agent collaboration in Agentic Web. We believe that this work will provide a standardized evaluation for software agent generation and will contribute to the future of scaling the capacity of Agentic Web.

cs.MA↗
arXiv2026

B-field Orion Protostellar Survey (BOPS). IV: The Relative Orientation Between Magnetic Fields and Density Structures in Young Protostellar Envelopes

We investigate the relative alignment between density structures and magnetic fields in eight young protostars from the ALMA B-field Orion Protostellar Survey. Column density maps are derived from 870 $\mu$m dust continuum emission, and the Histogram of Relative Orientations (HRO) method is applied to quantify the correlation between magnetic field orientations and density structures on envelope scales ($\sim$10$^{3}$~au). We find that the relative alignment shows overall weak evidence of systematic evolution with column density, suggesting that column density alone does not fully determine the alignment. The magnetization level also plays a crucial role, with weakly magnetized envelopes exhibiting predominantly parallel or random alignment, whereas strongly magnetized ones show perpendicular configurations even at moderate densities. These results reveal that density and magnetization jointly shape the morphology of protostellar envelopes and the coupling between gravity and magnetic fields during early stages of star formation.

astro-ph.GA↗
arXiv2026

StealthMark: Harmless and Stealthy Ownership Verification for Medical Segmentation via Uncertainty-Guided Backdoors

Annotating medical data for training AI models is often costly and limited due to the shortage of specialists with relevant clinical expertise. This challenge is further compounded by privacy and ethical concerns associated with sensitive patient information. As a result, well-trained medical segmentation models on private datasets constitute valuable intellectual property requiring robust protection mechanisms. Existing model protection techniques primarily focus on classification and generative tasks, while segmentation models-crucial to medical image analysis-remain largely underexplored. In this paper, we propose a novel, stealthy, and harmless method, StealthMark, for verifying the ownership of medical segmentation models under black-box conditions. Our approach subtly modulates model uncertainty without altering the final segmentation outputs, thereby preserving the model's performance. To enable ownership verification, we incorporate model-agnostic explanation methods, e.g. LIME, to extract feature attributions from the model outputs. Under specific triggering conditions, these explanations reveal a distinct and verifiable watermark. We further design the watermark as a QR code to facilitate robust and recognizable ownership claims. We conducted extensive experiments across four medical imaging datasets and five mainstream segmentation models. The results demonstrate the effectiveness, stealthiness, and harmlessness of our method on the original model's segmentation performance. For example, when applied to the SAM model, StealthMark consistently achieved ASR above 95% across various datasets while maintaining less than a 1% drop in Dice and AUC scores, significantly outperforming backdoor-based watermarking methods and highlighting its strong potential for practical deployment. Our implementation code is made available at: https://github.com/Qinkaiyu/StealthMark.

cs.CV↗
arXiv2026

A compact object with a K type star companion in the solar neighborhood: a wide post common envelope binary with a white dwarf candidate

Post-common envelope binaries (PCEBs) consisting of a white dwarf (WD) plus a main-sequence (MS) star can constrain current prescriptions of common envelope evolution (CEE) and calibrate theoretical models of binary formation and evolution. Most PCEBs studied to date have typical orbital periods of hours to a few days and can be well explained by assuming inefficient CEE to expel the envelope. However, there are currently several systems with relatively wide orbital periods ($>$18 days). To explain these wide PCEBs, additional sources of energy have been suggested to be taken into account. Here, we present the discovery and observational characterization of a compact object ($M\,\geq\,0.58\,\rm M_{\odot}$) with a K-type star companion in the solar neighborhood ($d\sim 112$ pc) and an orbital period of $P_{\rm orb}\sim 14$ days. The compact object binary is likely to be a system consisting of a WD and a barium dwarf, making it the shortest-period barium star binary candidate. Such a system with an orbital period within the gap between tight and wide binaries provides a test of whether additional energy sources are required to explain its formation. Using binary evolution models, we investigate the evolutionary history of this wide PCEB system and find that the observed properties of this source can be explained without invoking any extra energy source.

astro-ph.SR↗
arXiv2026

Holos: A Web-Scale LLM-Based Multi-Agent System for the Agentic Web

As large language models (LLM)-driven agents transition from isolated task solvers to persistent digital entities, the emergence of the Agentic Web, an ecosystem where heterogeneous agents autonomously interact and co-evolve, marks a pivotal shift toward Artificial General Intelligence (AGI). However, LLM-based multi-agent systems (LaMAS) are hindered by open-world issues such as scaling friction, coordination breakdown, and value dissipation. To address these challenges, we introduce Holos, a web-scale LaMAS architected for long-term ecological persistence. Holos adopts a five-layer architecture, with core modules primarily featuring the Nuwa engine for high-efficiency agent generation and hosting, a market-driven Orchestrator for resilient coordination, and an endogenous value cycle to achieve incentive compatibility. By bridging the gap between micro-level collaboration and macro-scale emergence, Holos hopes to lay the foundation for the next generation of the self-organizing and continuously evolving Agentic Web. We have publicly released Holos (accessible at https://holosai.io), providing a resource for the community and a testbed for future research in large-scale agentic ecosystems.

cs.AI↗
arXiv2026

Parallel Dynamic Spatial Indexes

Maintaining spatial data (points in two or three dimensions) is crucial and has a wide range of applications, such as graphics, GIS, and robotics. To handle spatial data, many data structures, called spatial indexes, have been proposed, e.g. kd-trees, oct/quadtrees (also called Orth-trees), R-trees, and bounding volume hierarchies (BVHs). In real-world applications, spatial datasets tend to be highly dynamic, requiring batch updates of points with low latency. This calls for efficient parallel batch updates on spatial indexes. Unfortunately, there is very little work that achieves this. In this paper, we systematically study parallel spatial indexes, with a special focus on achieving high-performance update performance for highly dynamic workloads. We select two types of spatial indexes that are considered optimized for low-latency updates: Orth-tree and R-tree/BVH. We propose two data structures: the P-Orth tree, a parallel Orth-tree, and the SPaC-tree family, a parallel R-tree/BVH. Both the P-Orth tree and the SPaC-tree deliver superior performance in batch updates compared to existing parallel kd-trees and Orth-trees, while preserving better or competitive query performance relative to their corresponding Orth-tree and R-tree counterparts. We also present comprehensive experiments comparing the performance of various parallel spatial indexes and share our findings at the end of the paper.

cs.DB↗
arXiv2025

TrimTokenator-LC: Towards Adaptive Visual Token Pruning for Large Multimodal Models with Long Contexts

Large Multimodal Models (LMMs) have proven effective on various tasks. They typically encode visual inputs into Original Model sequences of tokens, which are then concatenated with textual tokens and jointly processed by the language model. However, the growing number of visual tokens greatly increases inference cost. Visual token pruning has emerged as a promising solution. However, existing methods often overlook scenarios involving long context inputs with multiple images. In this paper, we analyze the challenges of visual token pruning in long context, multi-image settings and introduce an adaptive pruning method tailored for such scenarios. We decompose redundancy into intra-image and inter-image components and quantify them through intra-image diversity and inter-image variation, which jointly guide dynamic budget allocation. Our approach consists of two stages. The intra-image stage allocates each image a content-aware token budget and greedily selects its most representative tokens. The inter-image stage performs global diversity filtering to form a candidate pool and then applies a Pareto selection procedure that balances diversity with text alignment. Extensive experiments show that our approach can reduce up to 80% of visual tokens while maintaining performance in long context settings.

cs.CV↗
arXiv2025

StarCraft+: Benchmarking Multi-agent Algorithms in Adversary Paradigm

Deep multi-agent reinforcement learning (MARL) algorithms are booming in the field of collaborative intelligence, and StarCraft multi-agent challenge (SMAC) is widely-used as the benchmark therein. However, imaginary opponents of MARL algorithms are practically configured and controlled in a fixed built-in AI mode, which causes less diversity and versatility in algorithm evaluation. To address this issue, in this work, we establish a multi-agent algorithm-vs-algorithm environment, named StarCraft II battle arena (SC2BA), to refresh the benchmarking of MARL algorithms in an adversary paradigm. Taking StarCraft as infrastructure, the SC2BA environment is specifically created for inter-algorithm adversary with the consideration of fairness, usability and customizability, and meantime an adversarial PyMARL (APyMARL) library is developed with easy-to-use interfaces/modules. Grounding in SC2BA, we benchmark those classic MARL algorithms in two types of adversarial modes: dual-algorithm paired adversary and multi-algorithm mixed adversary, where the former conducts the adversary of pairwise algorithms while the latter focuses on the adversary to multiple behaviors from a group of algorithms. The extensive benchmark experiments exhibit some thought-provoking observations/problems in the effectivity, sensibility and scalability of these completed algorithms. The SC2BA environment as well as reproduced experiments are released in \href{https://github.com/dooliu/SC2BA}{Github}, and we believe that this work could mark a new step for the MARL field in the coming years.

cs.AI↗
arXiv2025

Diffusion Reconstruction-based Data Likelihood Estimation for Core-Set Selection

Existing core-set selection methods predominantly rely on heuristic scoring signals such as training dynamics or model uncertainty, lacking explicit modeling of data likelihood. This omission may hinder the constructed subset from capturing subtle yet critical distributional structures that underpin effective model training. In this work, we propose a novel, theoretically grounded approach that leverages diffusion models to estimate data likelihood via reconstruction deviation induced by partial reverse denoising. Specifically, we establish a formal connection between reconstruction error and data likelihood, grounded in the Evidence Lower Bound (ELBO) of Markovian diffusion processes, thereby enabling a principled, distribution-aware scoring criterion for data selection. Complementarily, we introduce an efficient information-theoretic method to identify the optimal reconstruction timestep, ensuring that the deviation provides a reliable signal indicative of underlying data likelihood. Extensive experiments on ImageNet demonstrate that reconstruction deviation offers an effective scoring criterion, consistently outperforming existing baselines across selection ratios, and closely matching full-data training using only 50% of the data. Further analysis shows that the likelihood-informed nature of our score reveals informative insights in data selection, shedding light on the interplay between data distributional characteristics and model learning preferences.

cs.CV↗
arXiv2025

FullPart: Generating each 3D Part at Full Resolution

Part-based 3D generation holds great potential for various applications. Previous part generators that represent parts using implicit vector-set tokens often suffer from insufficient geometric details. Another line of work adopts an explicit voxel representation but shares a global voxel grid among all parts; this often causes small parts to occupy too few voxels, leading to degraded quality. In this paper, we propose FullPart, a novel framework that combines both implicit and explicit paradigms. It first derives the bounding box layout through an implicit box vector-set diffusion process, a task that implicit diffusion handles effectively since box tokens contain little geometric detail. Then, it generates detailed parts, each within its own fixed full-resolution voxel grid. Instead of sharing a global low-resolution space, each part in our method - even small ones - is generated at full resolution, enabling the synthesis of intricate details. We further introduce a center-point encoding strategy to address the misalignment issue when exchanging information between parts of different actual sizes, thereby maintaining global coherence. Moreover, to tackle the scarcity of reliable part data, we present PartVerse-XL, the largest human-annotated 3D part dataset to date with 40K objects and 320K parts. Extensive experiments demonstrate that FullPart achieves state-of-the-art results in 3D part generation. We will release all code, data, and model to benefit future research in 3D part generation.

cs.CV↗