arXiv subjects

Hao Li

Publications and source records attributed to Hao Li.

At least 55 records · Page 3Linked to original sources

arXiv2026

JoyAI-Voice 2.0: A Full-Continuous Autoregressive Speech Generation Model with Semantic-Acoustic Joint Representation

We present JoyAI-Voice~2.0, an end-to-end anthropomorphic speech generation model built upon a fully continuous, dual-encoder architecture. Raw speech is encoded into continuous latents and partitioned into patches. Each patch is decomposed by a semantic-acoustic dual encoder into a semantically purified representation and an acoustic representation, which are fused and jointly fed to a causal autoregressive Transformer for planning. The Transformer predicts the conditioning for the next patch, and a local diffusion Transformer renders its full latents for 48\,kHz synthesis. The model is trained with a joint flow-matching and stop-prediction objective, followed by supervised fine-tuning and reinforcement learning with DiffusionNFT to improve model performance. It achieves the lowest average word error rate of 2.51\% on Seed-TTS, a 14.9\% relative reduction over the strongest baseline, and state-of-the-art attribute fidelity on InstructTTSEval in both Chinese and English, leading on 5 of 10 perceptual dimensions with the highest overall score of 0.893 on MDVD-Eval.

cs.SD↗

arXiv2026

RSI-Router: Evolving Subtask-Level LLM Routing and Skills for Cost-Efficient Agents

Practical deployment of large language model (LLM) agents requires strong task performance at affordable inference cost. For long-horizon agentic tasks, this performance-cost trade-off can be improved through within-task large-small model collaboration, as smaller models can handle some stages even when they cannot solve the full task. In this paper, we introduce RSI-router, a routing framework that constructs subtask-level model assignments and model-specific skills through recursive self-improvement over accumulated experience. Each iteration consists of four stages: Subtask Mining derives subtask definitions and identification rules from training trajectories; Routing Strategy Evolution proposes and evaluates diverse model assignments; Model-Specific Skill Evolution compares routed and large-model-only trajectories to diagnose failures and develop reusable execution skills; and Pareto-Optimal Router Selection updates the Pareto population using historical and newly generated routers while retaining dominated routers as experience for subsequent evolution. Routing between DeepSeek-V4.1-Flash and Qwen3.5-9B, RSI-router consistently surpasses the DeepSeek-only baseline at roughly half the inference cost (48.3%) across five agentic benchmarks. In particular, on ALFWorld, ScienceWorld, and WebShop, it cuts inference cost by 74.7-82.2% while simultaneously improving performance; on Terminal-Bench 2.0, it achieves a 16.7% relative performance gain at 18.0% lower cost. Moreover, RSI-router establishes a stronger performance--cost Pareto frontier than 9 routing methods.

cs.AI↗

arXiv2026

The Right Lesson at the Right Step: Deriving Control Updates for Self-Evolving Agents

Self-evolving agents improve future behavior by reusing past experience, typically as global prompts, memories, or reflections. Yet these mechanisms rarely control where experience takes effect. In long tool-use workflows, the same lesson may correct one decision but distract another, making experience reuse a problem of localized control rather than memory alone. We introduce EvoCUE (Evolution through Control Updates from Evidence), a framework for learning reusable control-program updates from completed agent executions. EvoCUE represents the agent as an explicit state-machine controller, whose nodes perform model or tool calls and whose edges define where control passes next. This makes the workflow editable at precise locations, so each learned update can specify what to add, where it acts, and when it applies. From completed trajectories, EvoCUE uses residual goals and observed execution traces to propose localized instruction or skill edits. Each candidate is evaluated at the point where it would act by resuming the parent and edited controllers from the same checkpoint and comparing their final outcomes. Accepted edits are compiled with applicability rules, confirmed on held-out tasks, and inherited by later executions. We evaluate EvoCUE on long tool-use environments where learned conventions must reach the right execution step. From a minimal AppWorld controller without benchmark-specific onboarding instructions, EvoCUE learns the missing task-completion convention and substantially improves success on Test-Normal and Test-Challenge. On PAST-Bench office workflows, EvoCUE transfers organizational requirements from prior episodes to later tasks, improving task-execution quality. These results show that self-evolving agents should place experience inside the control flow, rather than only store it as text.

cs.CL↗

arXiv2026

Decoding the Disaster: Multi-Task Geospatial Reasoning with Vision-Language Models and Crowdsourced Imagery for Disaster Mapping

Crowdsourced imagery provides timely, fine-grained, street-level observations for disaster mapping, complementing conventional remote sensing imagery (RSI) during emergency response. However, such imagery is often unstructured, spatially ambiguous, and lacks reliable geographic metadata, making manual geolocalization and interpretation labor-intensive and difficult to scale. This work proposes a multi-task Geospatial Reasoning Disaster mapping framework, namely GRDisaster, to examine the potential of vision-language models (VLMs) in understanding, geolocalizing, and reasoning over crowdsourced disaster imagery. GRDisaster is built on a newly curated benchmark dataset derived from PhotoMappers, comprising 26,340 images organized into human-validated volunteered geographic information (VGI), street-view imagery (SVI), RSI cross-view triplets covering multiple disaster events from 2018 to 2024. The framework combines deterministic and probabilistic cross-view geolocalization with multi-view fusion to associate VGI images with georeferenced SVI and RSI. It introduces two sets of spatial reasoning indicators for cross-view geolocalization validation and disaster damage assessment. These indicators use structural, environmental, and global-scene cues to validate cross-view correspondences and visually observable damage evidence with expert-verified annotations to assess disaster severity, improving the interpretability of VLM outputs. To our knowledge, this study provides the first systematic investigation and unified evaluation framework for examining how VLM-based spatial reasoning can transform crowdsourced disaster imagery into actionable geospatial artificial intelligence (GeoAI) through cross-view geolocalization validation, interpretable spatial reasoning, and damage-aware severity assessment.

cs.CV↗

arXiv2026

Learning to Generate Rigid Body Interactions with Video Diffusion Models

Recent video generation models have achieved remarkable progress and are now deployed in film, social media production, and advertising. Beyond their creative potential, such models also hold promise as world simulators for robotics and embodied decision making. Despite strong advances, current approaches still struggle to generate physically plausible object interactions and lack object-level control mechanisms. To address these limitations, we introduce KineMask, an approach for video generation that enables realistic rigid body control, interactions, and effects. Given a single image and a specified object velocity, our method generates videos with inferred motions and future object interactions. We propose a two-stage training strategy that gradually removes future motion supervision via object masks. Using this strategy we train video diffusion models (VDMs) on synthetic scenes of simple interactions and demonstrate significant improvements and generalization to rigid body and hand-object interactions in real scenes. Furthermore, KineMask integrates low-level motion control with high-level textual conditioning via predicted scene descriptions, leading to support for synthesis of complex dynamical phenomena. Our experiments show that KineMask generalizes to different VDMs and achieves strong improvements over recent models of comparable size. Ablation studies further highlight the complementary roles of low- and high-level conditioning in VDMs.

cs.CV↗

arXiv2026

Preserve, Reveal, Expand: Towards Faithful 4D Video Editing with Region-Aware Conditioning and Benchmarking

While 4D-driven video diffusion models can generate visually plausible videos, faithful 4D editing poses a stricter requirement: source-observed content should be preserved, while disoccluded and out-of-view regions must be synthesized naturally. We identify a key failure mode in existing methods, which we term Evidence-Role Mismatch: source-backed observations, uncertain rendered cues, and unsupported regions are entangled within a single conditioning signal, which can lead to preservation drift, ghosting, and unstable extrapolation. To mitigate this issue, we propose PREX (Preserve, Reveal, Expand), a region-aware framework that decomposes the target spatiotemporal volume into three distinct roles based on observation support and scene extent. PREX traces edited 4D points back to source frames and retrieves observed colors and pairs these appearance cues with geometry-aware confidence and region roles, and injects them into a frozen video diffusion backbone through an adapter trained with self-supervised proxy editing tasks. We further introduce PREBench, a diagnostic benchmark specifically designed for 4D video editing, to expose and quantify such region-specific failures. PREBench separately evaluates Preserve fidelity, Reveal quality, and Expand plausibility of videos generated by 4D driven video diffusion models. Experiments show that PREX substantially reduces region-specific failure modes while maintaining strong visual quality and 4D edit control compared with existing 4D-based video diffusion methods. Code will be publicly released. Project Page: https://ricepastem.github.io/PREX-Open

cs.CV↗

arXiv2026

Dream-Butterfly: Configuring Embodied Conversational Guidance for Public Outdoor Mixed Reality Exhibitions

Public outdoor mixed reality (MR) exhibitions make guidance a spatial interaction problem: visitors need timely explanations while roaming open sites, attending to virtual artworks, hazards, bystanders, and route choices. We present Dream-Butterfly, a field-deployed embodied conversational guide for campus-scale outdoor MR exhibitions. Visitors explicitly summon the lightweight non-humanoid guide; it returns visibly to a hand-relative dialogue position and answers using retrieval-augmented responses scoped by the MR runtime to the artwork currently encountered. Within the resulting mixed guidance ecology, staff remain accountable for safety, wayfinding, device support, and contingencies. Dream-Butterfly was deployed with over 30 spatially anchored artworks across a 26,000 m$^2$ campus site, which we use as a stress case for public MR guidance under walking, field-of-view limits, route choice, bystander exposure, and safety constraints. In an in-the-wild study with 24 visitors, we compared two role arrangements: agent-default, where the embodied guide served as the default channel for artwork interpretation, and docent-default, where staff provided primary narration while the agent remained optional. Agent-default guidance was associated with more available, visitor-timed explanations, higher immersion/engagement, and stronger hedonic quality, while increasing visitors' self-curation burden around pacing, attention management, and question formulation. We argue that making interpretation available at visitors' chosen moments improves the timing of explanation in outdoor MR and shifts more work to visitors: deciding when to stop, what to ask, how deeply to engage, and when to move on.

cs.HC↗

arXiv2026

Strangeness Production at GlueX

The GlueX experiment at Jefferson Lab is well suited for exclusive studies of strangeness production, combining a linearly polarized photon beam at $8.0 < E_γ< 8.8$ GeV with a large-acceptance detector that reconstructs weakly decaying hyperons. We outline a research program built on three aspects: (i) spectroscopy of strange mesons, where many excited $K^*$ states remain unestablished and crypto-exotic strange hybrids must be identified from combined spectrum and decay information, anchored by the $K^*(892)$ as a standard candle; (ii) spin dynamics, where the first precise measurements of the complete spin-density matrix elements in the photoproduced $K^*(892)Λ$ system beyond the Schilling formalism constrain the $t$-channel strange production mechanism; and (iii) hadronization, probed through the newly published $p\bar{p}$, $Λ\barΛ$, and $p\barΛ$ photoproduction cross sections and emerging $Λ\barΛ$ spin-correlation measurements. We place these efforts in the context of the broader GlueX strangeness program and give an outlook toward the high-statistics GlueX-II era.

hep-ex↗

arXiv2026

ELUCID. IX. Recovering the Substructures and History of the Coma Cluster

We employ constrained hydrodynamic simulations of the Coma galaxy cluster from the ELUCID project to study its substructures and assembly history. Our simulations accurately reproduce the global properties of Coma, including its position, virial mass, radius, and surrounding large-scale filaments. Using a combined HBT+SKID method, we obtain a total intracluster light (ICL) fraction of $12.2\%-23.5\%$, consistent with recent observations. The mass-weighted ICL profiles of velocity dispersion, stellar age, and iron abundance generally agree with MaNGA measurements, and its clumpy east-west elongation is closely linked to the merger history of the two brightest cluster galaxies (BCGs). The simulation also successfully reproduces the north and west intracluster filaments (ICFs) detected by weak lensing, and the surrounding galaxy groups are distributed in alignment with the directions toward the nearby A2199 and A1367 clusters. The simulation further predict a complex cluster assembly history, with two major mergers at $z=0.74$ and $0.45$, and a pericentric passage of the two BCGs occurring $\sim 0.64 \ {\rm Gyr}$ before its current state. Our results demonstrate that constrained simulations are a powerful tool for connecting observed structures to the unobservable assembly histories of individual galaxy clusters.

astro-ph.GA↗

arXiv2026

Noise-Robust Quantum State Characterization for Remote State Preparation with Deep Learning

Quantum communication underpins secure information processing and scalable quantum networks. In particular, remote state preparation (RSP) enables efficient quantum state transfer, but accurately estimating target states under complex noise remains challenging. Here, we propose a Transformer-based Quantum State Characterizer (TQSC) model for noisy RSP experiments. Our model reconstructs experimentally prepared pure and mixed photonic polarization states from noisy measurements in complex scattering environments, while its attention patterns provide physically grounded insights into correlations among the measured observables. The method achieves a mean estimator-target fidelity exceeding 99.999% under complex scattering and dynamic Gaussian noise, while its robustness and generalization are further examined using Qiskit-simulated Bloch-ball states. Furthermore, in a practical MNIST image transmission task with held-out states, the decoded bit error rate is reduced from 50.34% to zero after TQSC post-processing. The TQSC model enables accurate tomographic characterization under dynamic noise and provides physically grounded post-hoc insights, holding promise for intelligent quantum information processing applications.

quant-ph↗

arXiv2026

Testing the strong equivalence principle with multimessenger binary neutron star mergers

The constancy of the gravitational constant $G$ is a cornerstone of the strong equivalence principle and of general relativity, yet its possible temporal variation remains a key target in tests of fundamental physics. Gravitational-wave (GW) astronomy, especially when combined with electromagnetic observations, provides an unprecedented new opportunity to probe this principle in the strong-field and dynamical regime. In this work, we develop a GW waveform model with a slowly varying gravitational constant, incorporating its effects both on compact binary dynamics and GW propagation in an expanding universe. Applying this framework to the binary neutron star merger GW170817, together with independent electromagnetic constraints on the luminosity distance, sky localization and binary inclination from GRB 170817A, we perform a joint Bayesian analysis that disentangles varying-$G$ effects from astrophysical degeneracies. We find no evidence for a temporal variation of the gravitational constant, and constrain its fractional time derivative to $\dot{G}/G \in [-3.36 \times 10^{-9}, 5.34\times10^{-10}]~{\rm yr^{-1}}$, representing the most stringent bounds obtained to date from real GW observations. Our results demonstrate the power of multi-messenger astronomy as a precision probe of the strong equivalence principle in the relativistic regime.

gr-qc↗

arXiv2026

DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression

The widespread adoption of long-horizon agents has made model workloads increasingly input-heavy. Although prior work has substantially reduced the cost of long-context computation, prefill remains computationally expensive, and large KV caches continue to strain HBM and SSD capacity and data-transfer bandwidth. Together, these compute, storage, and bandwidth demands constitute the primary bottleneck to further lowering deployment costs. To address this challenge, we introduce DeepSeek-V4.1-Flash, a multimodal Mixture-of-Experts (MoE) model with 552B backbone parameters and support for contexts of up to one million tokens. With its Causal Encoder-Decoder (CED) architecture, the model activates 16B parameters per token during decode but only 8B parameters during prefill, substantially improving cost efficiency for agentic workloads. To push the limits of KV cache compression, DeepSeek-V4.1-Flash combines cross-layer KV cache reuse in Compressed Sparse Attention 2 (CSA2) with FP4 KV caching. These designs reduce its global KV cache footprint (always in HBM) to 890 bytes per token, roughly 1/4 of the corresponding footprint of DeepSeek-V4-Flash. Further, through a dedicated deployment optimization known as SWA Bounded Replay, DeepSeek-V4.1-Flash reduces its persistent KV cache footprint (always on SSD or in host memory) to roughly 1/8 of that of DeepSeek-V4-Flash. Despite its much smaller KV cache footprint, the model delivers substantially better performance than the baseline. In addition, we streamline the DeepSeek-V4 architecture and introduce several efficient architectural extensions. We pretrain DeepSeek-V4.1-Flash on a multimodal corpus comprising 45T tokens and conduct comprehensive post-training, yielding strong performance across diverse text-based and multimodal agentic scenarios. Model checkpoints are available at https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash.

cs.CL↗

arXiv2026

Meta-LinEXP3: Online-within-Online Learning for Adversarial Linear Contextual Bandits

Meta-learning has emerged as an effective paradigm for transferring knowledge across sequential bandit tasks. While substantial progress has been made for stochastic bandits and non-contextual adversarial bandits, meta-learning for adversarial linear contextual bandits (ALCBs) with random action sets remains largely unexplored. To address this problem, we propose Meta-LinEXP3, an online-within-online algorithm that constructs a predictable task-level prior from completed tasks to guide the inner LinEXP3 learner. For known context distributions, we develop a policy-centered estimator that achieves an intrinsic-dimension $\mathcal{O}(\sqrt{n})$ per-task regret bound. For unknown distributions, we introduce a past-only regularized moment estimator with an $\mathcal{O}(n^{2/3})$ leading regret term and explicit finite-sample error. We further establish a direct connection between prior accuracy and transfer regret, showing that increasingly accurate priors yield sublinear transfer-dependent regret across tasks. Experiments demonstrate the effectiveness of Meta-LinEXP3, including its application to structured hyperspectral tensor sampling.

cs.LG↗

arXiv2026

Spectral-Residual Continuous Greedy for Tensor Sampling

Sampling a multidomain tensor from limited measurements is fundamental in structured linear inverse problems. Kronecker-structured sampling avoids the full high-dimensional sensing matrix, but design remains difficult: sequential discrete methods can commit the cross-mode budget too early, whereas standard continuous greedy avoids such early commitment at the cost of repeated state-dependent gradient evaluations. We propose spectral-residual continuous greedy (\alg) for frame-potential (FP) tensor sampling. \alg maintains a state-dependent fractional allocation before rounding. Mode-wise Gram matrices provide safe gradient intervals for direction certification, while Shapley values prioritize unresolved gradient queries. SR-CG then applies deterministic rounding followed by path-guided exchange (PGX), which reuses the final fractional state to restrict candidate swaps and accepts only exact FP-decreasing exchanges. We establish a finite-step approximation guarantee that approaches the classical $1-1/e$ factor as direction certification becomes exact and finite-step residual error vanishes. Experiments show fewer exact gradient evaluations and better FP designs, with clearer gains on instances where the cross-mode budget allocation is difficult to determine. \alg also achieves lower average normalized mean-squared error (NMSE) than Greedy-FP at all tested noise levels, although the reconstruction gain is smaller than the FP gain.

math.OC↗

arXiv2026

AI for Games in the Foundation Model Era

Foundation models, alongside advances in learned game-world models, are reshaping AI across the game lifecycle. Beyond playing games, recent systems model players and game dynamics, support design and development, adapt player-facing experiences at runtime, and evaluate resulting artifacts. Yet these directions have evolved largely separately, obscuring which capabilities transfer across settings and which remain tied to particular games, engines, interfaces, or player populations. We organize the literature into six roles according to the immediate use of AI output: playing and acting; modeling players and games; designing games; building and maintaining games; generating and adapting at runtime; and testing and evaluating games. For each role, we examine what structure is supplied by the game or workflow, what AI learns or produces, which capabilities and artifacts transfer across settings and roles, and what evidence supports the claims. We identify cross-role connections: trajectories train world models, learned environments provide experience for agents, design specifications drive executable implementations, and play or testing feedback guides revision. However, control schemes, rules, engine interfaces, state representations, and player contexts often remain setting-specific, so downstream claims require validation in the target setting. Evaluation is most standardized for bounded game playing and selected learned environments, while persistent state in learned worlds, repeated software revision, validated player modeling, sustained runtime adaptation, and representative automated testing remain less established. The central challenge is to reuse or transfer outputs and capabilities across roles while re-establishing evidence for effectiveness in the game-specific contexts where they are used.

cs.AI↗

arXiv2026

Beyond the Previous Layer: Residual Predictive Structure in Sparse MoE Routing

Sparse mixture-of-experts models route each token through a sequence of expert selections. We ask whether the immediately preceding selection adequately summarizes this trajectory for predicting the next router. Using frozen OLMoE and JetMoE models, we measure the held-out predictive gain from earlier expert selections while retaining the most recent selection as a common baseline. In OLMoE, extending the history from one to eleven layers raises router-logit $R^2$ from 0.59879 to 0.66544. A preregistered JetMoE replication yields four-layer gains of 0.14275 and 0.20528 at two target depths, with paired bootstrap intervals above zero. These gains survive nonlinear decoding: adding history to a small multilayer perceptron improves $R^2$ by 0.17137 and 0.21861, whereas nonlinear decoding of the recent state alone adds 0.00139 and 0.00936 over a linear probe. Parameter-matched controls preserve the advantage, and cross-fitted history residuals predict target residuals with $R^2$ of 0.20549 and 0.23556. These findings identify residual predictive structure in expert-selection trajectories beyond adjacent-layer persistence.

cs.LG↗

arXiv2026

The Fixed Server Locality Gap of Count Load Assignment Games

Marginal-contribution pricing makes the social objective an exact potential, but does not ensure that every stable assignment is efficient. We study indivisible clients with heterogeneous workloads, arbitrary nonnegative assignment costs, and private eligibility menus over a fixed number of shared servers and optional local execution. Each server's social cost is its occupancy multiplied by its aggregate workload and a server-specific coefficient. For every fixed server cap $M$, we prove that the price of anarchy over this class is $Θ_M(κ^{1-2^{-M}})$ as the workload-ratio cap $κ$ grows. The upper bound applies to every equilibrium and every feasible comparison, without an acyclic-comparison or strongly-connected-component restriction. Its proof combines a common-multiplier certificate, a source-weighted residual inequality and an ordered moment recurrence. A matching chain family has a realization by mobile unmanned aerial vehicle (UAV) clients and fixed ground servers, with positive altitude, positive-width coverage and an explicitly paid wireless baseline. We prove exact waypoint elimination and give an input-computable component refinement. Scheduling and congestion identities delimit which established bounds transfer. Exact enumeration of 2,800 synthetic games identifies inefficient equilibria, while 270 primary-parameterized synthetic instance-budget records evaluate incremental latency relative to a feasible multistart reference. Tightness concerns the heterogeneity exponent for fixed $M$, not matching leading constants or measured frequency of the worst-case configurations.

cs.GT↗

arXiv2026

OpenAI4S: Code as Action, Science as Sessions

AI co-scientists could accelerate computational research, but over a long-running study the workflow also has to stay inspectable, resumable and reproducible, which requires persistent computational state and provenance. Here we present OpenAI4S, an open-source scientific research agent built around the principle of \emph{Code as Action, Science as Sessions}. OpenAI4S combines a persistent computing runtime with research-session management: orchestration is handled through structured tool calls, while scientific actions are represented as complete code cells executed in persistent Python and R kernels. An append-only Action Ledger, per-cell execution records, versioned artifacts, environment records, and workspace checkpoints preserve how results were produced and support session recovery, branching, and extension. Configurable sandboxing, permission controls, and code and trajectory screening provide complementary safeguards. We evaluate OpenAI4S on 36 research scenarios spanning retrosynthesis, molecular dynamics, protein binder design, protein mutation, catalyst screening, and mineral spectroscopy, measuring scientific task accuracy, workflow completeness, and reproducibility of the resulting repositories. OpenAI4S achieves an overall score of 7.83, compared with 5.7--6.4 for a general-purpose coding harness evaluated with three frontier models, with the largest gains on long-horizon and computation-intensive workflows. These results suggest that integrating persistent execution with session-level provenance can improve the reliability of AI-assisted scientific workflows. Environment specification and full rerunnability remain weak for every evaluated system, ours included, so reproducibility is still an open problem for scientific agents. The system is available under the MIT license at \href{https://github.com/PKU-YuanGroup/OpenAI4S}{github.com/PKU-YuanGroup/OpenAI4S}.

cs.AI↗