Searcharxiv⌕ Search

arXiv subjects

Hao Guo

Publications and source records attributed to Hao Guo.

At least 19 recordsLinked to original sources
arXiv2026

MLLM-GS: Multimodal Large Language Model Prior-Guided Fast Novel View Synthesis in Extreme Environments

Novel view synthesis (NVS) in degraded scenes aims to reconstruct 3D scenes from multi-view observations affected by various degradations, and to generate clear and photorealistic images at unseen viewpoints. Despite recent advances, existing methods often rely on degradation-specific models, limiting their generalization across diverse extreme environments. Generative models provide rich visual priors, but their outputs often suffer from pixel misalignment and color deviations, which can propagate into the reconstructed 3D scene and compromise reconstruction quality. Furthermore, the high computational and memory costs of 3D reconstruction hinder the practical application of NVS in extreme environments. To this end, we propose MLLM-GS, a multimodal large language model (MLLM) prior-guided framework for fast NVS in extreme environments. MLLM-GS consists of two branches: MLLM prior-guided view enhancement and fast 3D scene reconstruction. The former leverages the rich visual priors of a frozen MLLM to enhance views under different severe degradations and introduces a lightweight correction network to reduce pixel shifts and color distortions in MLLM outputs. With supervision from only a few paired clean views, this branch produces reliable enhanced views for all degraded observations, providing pseudo-GT views for subsequent 3D reconstruction. The latter improves 3DGS efficiency by optimizing rasterization, Gaussian memory layout, and parameter updates, reducing redundant computation, memory-access overhead, and optimization costs. Extensive experiments on multiple degraded scenes demonstrate that MLLM-GS effectively handles various severe degradations and achieves high-quality NVS with significantly lower reconstruction time and memory consumption. Project Page: https://wzlwlm13.github.io/MLLM-GS

cs.CV↗
arXiv2026

Quantitative $K$-homology of uniform Roe algebras

We introduce quantitative $K$-homology for uniform Roe algebras of discrete proper metric spaces with bounded geometry, based on the controlled $KK$-theory of Willett and Yu. We construct a comparison map from Lipschitz $K$-theory and prove that, for chain geodesic spaces, it is an isomorphism in a quantitative sense; in particular, quantitative $K$-homology is isomorphic to the inverse limit of the Lipschitz $K$-theory groups as the Lipschitz constant tends to zero. We also establish a pairing between quantitative $K$-homology and quantitative $K$-theory.

math.KT↗
arXiv2026

Weak-$L^\infty$ and BMO with Hausdorff Content

The space of weak-$L^{\infty}$ on measure space was introduced by Bennett et al. (Ann. of Math., 113(1981), 601-611) and proved to contain the space of $L^\infty$ and possess the appropriate interpolation property. In this paper, we are aimed to continue this line of research on the Euclidean space $\mathbb R^n$ endowed with Hausdorff content, which is a low-dimensional outer measure on $\mathbb{R}^n$, and fails to be additive or absolutely continuous with respect to Lebesgue measures. But low-dimensional subsets, such as fractals, of $\mathbb{R}^n$ are Hausdorff capacitable. Precisely, in the setting of Hausdorff content, an analogue of weak-$L^\infty$ is defined and proved to possess an improved Marcinkiewicz interpolation theorem. Moreover, we show that the space of functions with bounded mean oscillation associated to the Hausdorff content is embedded into this new weak-$L^\infty$ space by establishing a key estimate for the distribution via a packing condition. As a by-product, the boundedness of the Hardy-Littlewood operator on the space of weak-$L^{\infty}$ is established.

math.CA↗
arXiv2026

PAFS: Perturbation-Aware Firing Scaffolds for Robust Event-Based SNNs

Event-based spiking neural networks (SNNs) encode information through temporal firing dynamics, but the same stateful computation makes their predictions sensitive to event loss, timestamp jitter, and speed variation. These perturbations alter evidence accumulation, yet late-layer firing trajectories can retain error-diagnostic structure that is absent from the final logits alone. We introduce Perturbation-Aware Firing Scaffolds (PAFS), a labeled post-training calibration method that exploits this structure without updating the pretrained backbone. PAFS associates standardized late-layer leaky integrate-and-fire trajectories with the class-probability residuals of an Adapter anchor, then applies singular value decomposition to extract compact residual-aligned channel directions. At inference, firing trajectories from the same backbone pass are projected onto this fixed low-rank scaffold and combined with anchor statistics to predict a residual logit correction. Across four matched dataset-backbone settings spanning three event datasets and two spiking Transformer architectures, PAFS improves corrupted-average and worst-condition accuracy over the corresponding Adapter anchors, with maximum gains of 3.06 and 3.65 percentage points. In the DailyAction-DVS/Spikingformer setting, the added trainable parameters amount to approximately 0.34% of the frozen backbone.

cs.NE↗
arXiv2026

Nemotron 3.5 Bolt 74B-A9B 24GB: A Hybrid Vision-Language Model for Local AI

We introduce Nemotron 3.5 Bolt 74B-A9B 24GB, a hybrid Mamba-Transformer vision-language model that optimizes speed and intelligence per GB of memory used for local agentic workloads. Derived from the Nemotron 3.5 Super VL 120B-A12B model family, Nemotron 3.5 Bolt 74B-A9B 24GB combines Star Elastic compression with 2-bit Quantization-Aware Distillation for memory-constrained deployment. The model is optimized for reasoning, agentic tool use, and web & terminal interactions. We also trained a LiLiCorr speculative drafter that is released alongside the model to reach our target inference speed. The resulting model achieves a higher accuracy than leading models with similar memory footprints and speed across many benchmarks for general agentic capabilities (e.g., GDPVal v2), web browsing (e.g., BrowseComp), and terminal use (e.g., Terminal-Bench 2.1), with a relatively high TPS of 84 on RTX Spark that is comparable to models with much lower active parameter counts and more than sufficient for long-horizon agentic applications.

cs.CV↗
arXiv2026

Metric-Changing Quasi-Hermitian Quenches: Normalized Loschmidt Amplitude and Complex-Time Zeros

In quasi-Hermitian mixed-state quenches, the initial and final Hamiltonians generally possess different positive-definite metrics, so post-quench evolution is not quasi-unitary in the initial physical space. Choosing the preparation/measurement space as reference, we derive from first principles a normalized mixed-state Loschmidt amplitude using metric-compatible purifications. The resulting biorthogonal trace representation contains a normalization factor given by the norm of the evolved purification in the initial metric; this factor is essential for comparing states from different physical Hilbert spaces and directly encodes the metric mismatch. We prove that this normalization function is strictly positive on the real-time axis, so real-time criticality arises solely from destructive interference in the dynamical overlap. In complex time, zeros of the normalization factor generate branch singularities distinct from Fisher zeros. For a general two-level system, we derive analytic expressions for the Fisher and normalization zeros and analyze exceptional-point scaling together with the temperature- and quasi-Hermiticity-dependent DQPT critical times in two solvable quench protocols. In contrast to metric-stationary quenches, where non-Hermitian effects may be absorbed into an equivalent Hermitian description, the metric mismatch here produces genuine non-Hermitian dynamics: a time-dependent normalization factor and complex-time branch structure that cannot be removed by any global similarity transformation.

quant-ph↗
arXiv2026

Higher localised $\widehat{A}$-genera: a note on disconnected fixed-point sets

For a proper cocompact action of a finitely generated discrete group $Γ$ on a spin manifold $M$, the higher localised $\widehat{A}$-genera $\widehat{A}_g(M,ω)$ introduced earlier by the author and Mathai were shown to be an obstruction to $Γ$-invariant metrics of positive scalar curvature, for $ω$ arising from the subring of $H^*(\underline{B}Γ,\mathbb{R})$ generated by elements of degree at most $2$, under the assumptions that the fixed-point set $M^g$ is connected and that $g$ is regular with respect to a certain real group $2$-cocycle. We show that both assumptions can be dropped.

math.DG↗
arXiv2026

Uhlmann Geometry of Fixed-Rank Density Matrices: A Fiber-Bundle Approach

For full-rank density matrices the state space is contractible, the Uhlmann bundle is topologically trivial, and the holonomy admits no quantized invariants; although the Uhlmann phase is genuinely geometric, this topological poverty has kept it from serving as a robust physical diagnostic of mixed-state matter. Mixed states of fixed rank below the Hilbert-space dimension, however, are ubiquitous: reduced states of constrained subsystems, states confined to invariant sectors, and states supported on decoherence-free subspaces can have a support smaller than the Hilbert space, a support that can vary with parameters. Their geometry is far richer: the support carries genuine curvature, non-Abelian holonomy, and Chern topology. We formulate Uhlmann's theory directly on the manifold of rank-$k$ density matrices: minimal purifications make the fixed-rank stratum the base of a principal $U(k)$-bundle whose Uhlmann connection is uniquely determined by a Sylvester equation, and in an eigenframe this connection takes a closed form reducing to the Berry, Wilczek--Zee, and faithful Uhlmann connections at $k=1$, at equal weights, and at $k=N$. The fixed-rank bundle inherits the topology of the Grassmannian, and non-factorizable higher Chern topology requires failure of the global eigenline splitting, which in the present spectral setting requires degeneracy, minimally realized by a rank-2 Yang monopole whose quantized second Chern number links the geometry to the four-dimensional quantum Hall response. Solvable models, from a genuinely non-Abelian holonomy to a dissipatively driven orbit whose Uhlmann holonomy is steered by the reservoir through an elliptic-integral spectral dressing, illustrate the content: the supporting subspace carries the topology, the spectral weights shape the transport, and Uhlmann holonomy provides a direct geometric probe of the resulting mixed-state structures.

quant-ph↗
arXiv2026

LiLiCorr: Lightweight Likelihood Correlation of Parallel Drafts for Speculative Decoding

Speculative decoding accelerates language-model inference by drafting future tokens the target model verifies in parallel. A diffusion-style drafter such as DFlash drafts an entire block in one forward pass. It is trained on the per-position marginals rather than on the joint distribution over the block, so the tokens it emits are individually plausible yet jointly incoherent. We introduce LiLiCorr, a Lightweight Likelihood-based model that Correlates the per-position marginals such a drafter produces. It keeps the top-K tokens at each position and processes them jointly, emitting an in and an out vector for each. Two candidates at consecutive positions match when the earlier out vector aligns, in cosine similarity, with the later in vector. Training scores the correct pairings highest and pushes competing ones down, so coherent blocks outscore incoherent ones. The joint distribution over the block, exponential in its length, is never materialized. One lightweight network pass produces all the vectors, the pairwise scores follow as batched matrix operations, leaving only a cheap greedy walk sequential. We co-train the DFlash drafter with LiLiCorr, so it proposes candidates that correlate into longer accepted sequences. Over the vanilla DFlash drafter it builds on, LiLiCorr accepts more and serves faster at all 72 settings we test: nine benchmarks at two target sizes under greedy and temperature-one decoding, plus a throughput sweep over six concurrencies, two input lengths and three output-entropy tiers. It raises acceptance length by 7 to 19%, while its single-pass scoring head costs only about 3% of the per-block latency. Against three concurrently developed methods that also restore coherence at draft time, all equally optimized on a common stack, LiLiCorr holds the highest throughput in 63 of those settings, ties within a measured noise floor in 6, and trails in only 3.

cs.CL↗
arXiv2026

FireWorldBench: Benchmarking Complex Physical World Intelligence through Coupled-Field Fire Dynamics

Understanding the physical world requires more than object recognition, scene description, and short-term visual prediction, as real-world physical systems involve multiple continuous fields, latent causal mechanisms, partial observations, and intervention-sensitive dynamics. We propose FireWorldBench, a benchmark for evaluating complex physical world intelligence in multimodal large language models and agents through coupled-field fire dynamics. Fire provides a canonical stress-test environment, where multiple interacting physical fields jointly shape observable states and temporal dynamics. FireWorldBench is organized along two complementary axes, a physical capability axis and a fire scenario task axis, jointly covering physical-state understanding, temporal dynamics, causal mechanisms, and intervention reasoning. The benchmark comprises 520 fire-world entries, including 494 controlled simulation worlds and 26 real-world-aligned event groups, spanning 47 scene archetypes across 7 environment families. These entries combine structured textual observations, multiple 2D physical-field visualizations, and 3D event-level scene modeling, yielding 9,074 text-image interleaved question-answer pairs across choice-based and open-ended report-generation formats. FireWorldBench evaluates whether models can infer latent physical states, explain underlying mechanisms, forecast coupled-field evolution, and assess intervention consequences from multimodal partial observations, providing a challenging testbed for complex physical world intelligence.

cs.AI↗
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

Torsion-induced gauge structure in curved quantum waveguides

We investigate the effective dynamics of a particle confined near a space curve. In the strict thin-layer reduction of the nondegenerate transverse ground state, torsion does not enter the local effective Hamiltonian, which contains only the curvature-induced scalar geometric potential. In contrast, for a thin guide with finite transverse width, a leading-order adiabatic projection onto the twofold-degenerate first-excited transverse band renders the rotation of the Frenet normal frame dynamically relevant and generates a matrix-valued Abelian gauge potential. Using a projection-based derivation in a co-rotating Frenet-frame basis, we show that this effective gauge potential is directly determined by the local torsion of the curve. The resulting effective Hamiltonian takes a gauge-covariant form and produces two transverse-mode branches whose parabolic dispersions are shifted in opposite directions in momentum space. For closed curves, the associated holonomy is controlled by the integrated torsion and leads to geometric interference. These results provide a direct realization of a Wilczek--Zee-type connection induced purely by spatial geometry in curved quantum waveguides. We further construct a classical-wave analogue using the degenerate bending modes of an isotropic elastic rod, demonstrating that the same torsion-induced gauge structure appears in continuum wave physics.

quant-ph↗
arXiv2026

SkillChain: Closing the Loop on Skill Evolution for Image-Based E-Commerce AI Assistants

Image-based AI assistants are now deployed at production scale on e-commerce platforms, where a single uploaded image can trigger fundamentally different user intents: product search, style recommendation, visual encyclopedia, or utility tool calls, each demanding its own response format, tool invocation, and domain knowledge. Without per-intent behavioral constraints, LLM-based systems conflate these heterogeneous modes and fall short of domain quality standards, while the breadth and dynamism of the intent space render manual engineering infeasible. To address this, we present SkillChain, which closes the production feedback loop on Skill evolution, automating the lifecycle of Skills through three stages: Skill Creator for bootstrapping from task specs and trajectories, Route Optimizer for routing alignment, and Body Refiner for iterative Skill Body refinement via dual-path LLM-Judge evaluation. Deployed on a production-scale e-commerce image assistant, SkillChain substantially improves aggregate response quality, with the strongest gains on structural compliance and content quality; a one-week online A/B experiment further confirms significant gains in user engagement, content consumption, and long-term retention.

cs.CL↗
arXiv2026

Dynamic Gated Cross-Modal Fusion with Sarcastic-aware Contrastive Regularization for Multimodal Sarcasm Detection

Multimodal sarcasm detection aims to identify sarcastic intent from multimodal content, where inconsistencies between literal meaning and contextual cues often signal irony. This task has attracted increasing research attention. However, accurate detection remains challenging due to instance-dependent modality contributions and misleading semantic consistency, where surface-level alignment masks underlying contradictory intent. Existing methods often rely on fixed fusion strategies and treat sarcasm as generic cross-modal mismatch, limiting their ability to capture subtle sarcasm cues and instance-specific modality interactions. To address these challenges, we propose a novel MSD framework that integrates Dynamic Gated Cross-Modal Fusion with Sarcastic-aware Contrastive Regularization (SaCR). Specifically, a bidirectional gated interaction module performs cross-modal feature filtering and adaptively calibrates textual and visual contributions at the instance level. A dynamic fusion gate further balances modality importance to generate more robust multimodal representations. Furthermore, SaCR is introduced as a label-aware contrastive regularization objective that encourages semantic consistency for non-sarcastic samples while suppressing misleading consistency in sarcastic cases. The proposed framework is trained end-to-end with a multi-objective learning strategy that jointly optimizes multimodal classification and auxiliary unimodal supervision. Extensive experiments on MMSD and MMSD2.0 demonstrate that the proposed method consistently outperforms strong baselines.

cs.CL↗
arXiv2026

Robust Incomplete Multimodal Sentiment Analysis via Iterative Proxy Correction

Multimodal sentiment analysis aims to infer affective states by integrating language, visual, and acoustic cues. However, real-world multimodal inputs are often incomplete or corrupted, which can weaken cross-modal complementarity and introduce misleading information into downstream fusion. Existing proxy-based methods for incomplete MSA commonly rely on one-shot proxy construction to compensate for degraded language information, but the generated proxy may be coarse or unreliable at initialization. Prematurely injecting such a proxy into multimodal reasoning can propagate initial errors and compromise sentiment prediction. To address this limitation, we propose an iterative proxy correction framework for robust incomplete MSA. Our method constructs a language-oriented proxy from non-language modalities and progressively refines it under multimodal context through gated residual correction. The corrected proxy is then adaptively fused with the observed language representation according to an estimated language reliability score, allowing the model to balance proxy-based compensation and trustworthy linguistic evidence. In addition, we introduce a stage-wise latent correction objective that uses the complete language representation as a training-time semantic anchor to stabilize the proxy refinement trajectory. Extensive experiments on MOSI, MOSEI, and SIMS under diverse missing-modality settings demonstrate that the proposed framework consistently outperforms competitive baselines and achieves robust sentiment prediction under incomplete inputs.

cs.CL↗
arXiv2026

Theory of the Uhlmann Phase in Quasi-Hermitian Quantum Systems

Geometric phases play a fundamental role in understanding the geometric structure of quantum states, yet extending the Uhlmann phase to non-Hermitian systems poses significant challenges due to parameter-dependent inner product structures. In this work, we develop a comprehensive theory of the Uhlmann phase for quasi-Hermitian systems, where the physical Hilbert space metric varies with external parameters. By constructing a generalized purification that respects the quasi-Hermitian inner product, we derive the corresponding parallel transport condition and Uhlmann connection. Our analysis reveals that the parameter-dependent metric modifies the Uhlmann connection and leads to a finite-temperature distribution of Uhlmann-phase regions that differs from the standard Hermitian case. Applying this formalism to solvable two-level models, we uncover tunable finite-temperature Uhlmann-phase diagrams, where the parameter-dependent metric shifts and deforms the zeros of the Uhlmann amplitude, thereby reshaping the temperature intervals in which nontrivial Uhlmann phases occur. Furthermore, by extending established interferometric protocols originally developed for Hermitian systems, the geometric amplitude can be recast as a measurable Loschmidt amplitude between purified states, providing a practical and experimentally accessible pathway to investigate quasi-Hermitian mixed-state geometric phases and their finite-temperature transitions. This work establishes a unified framework for understanding mixed-state geometric phases in quasi-Hermitian quantum systems and, through its natural relation to the mixed-state quantum geometric tensor, opens new avenues for exploring local geometry in quasi-Hermitian thermal states.

quant-ph↗
arXiv2026

A Frequency-Space Terahertz Transceiver Chip for Multi-Agent Communications and Spatial Awareness

Future indoor embodied-intelligence systems require scalable hardware platforms that support both high-capacity multi-agent connectivity and mutual spatial awareness. The terahertz (THz) spectrum offers abundant bandwidth and inherent spatial selectivity for integrated sensing and communication (ISAC); however, conventional phased arrays and programmable metasurfaces rely on dense beamforming networks, element-level control, or external THz illumination, making scalable multibeam operation challenging. Here, we report a fully integrated 208-258GHz 65-nm CMOS THz transceiver chip that monolithically integrates broadband front ends with heterogeneous leaky-wave metasurface (HLM) apertures within a 1.5mm by 4.9mm area. The HLM generates strongly dispersive leaky modes, enabling 75 degree frequency-controlled beam scanning with only four meta-atoms. Co-design of frequency-domain and spatial-domain mixing achieves spectrally clean frequency-to-space mapping for spatial-frequency division multiple access (SFDMA) communication. The THz chip demonstrates multi-agent simultaneous transmission and reception, two-dimensional localization, and sensing-enhanced communication, providing a scalable hardware platform for future THz embodied-intelligence networks.

physics.app-ph↗
arXiv2026

Freeform super-oscillatory optics for CMOS-integrated THz super-resolution imaging

The diffraction limit fundamentally constrains the spatial resolution of far-field imaging systems. While near-field techniques can circumvent this limit, their inherently short working distances (WD) severely restrict practical applications. Super-oscillatory lenses (SOLs) offer a far-field alternative; however, conventional SOLs are plagued by discrete operating wavelengths, low efficiencies (below 5%), and formidable trade-offs among numerical aperture, chromatic aberration, and depth of focus (DOF). Here, we introduce a nonlocal, nonlinear-curvature mechanism to design a freeform SOL that achieves ultrabroadband (0.3 to 1 THz), achromatic super-resolution focusing with an unprecedented efficiency of 44%. Operating at a 9 mm WD, the lens maintains a consistent sub-diffraction full-width at half-maximum (FWHM) of around 0.45 wavelength alongside an extended DOF of around 10 wavelengths. By integrating a compact 65-nm CMOS oscillator-radiator array, we establish an advanced imaging platform capable of resolving complex 2D and 3D sub-millimeter features (down to 0.15 mm). Readily scalable to the optical regime via two-photon lithography, this freeform SOL paradigm paves the way for next-generation, high-performance integrated photonics.

physics.optics↗