arXiv subjects

Pengfei Zhu

Publications and source records attributed to Pengfei Zhu.

At least 19 recordsLinked to original sources

arXiv2026

A State-Variable Theory of Evolving Material Symmetry

Material architecture can evolve during loading, yet its mechanically expressed symmetry depends on both the structural state and the constitutive coupling. We connect these quantities through one parent-invariant free energy with a symmetric traceless internal tensor $\mathbf Q$. Its stabilizer distinguishes isotropic, uniaxial, and biaxial states, and projected gradient kinetics satisfies the structural dissipation inequality. Parent invariance generally guarantees only symmetry inheritance. For an explicit two-coupling elastic potential, we derive a stiffness-contraction reconstruction of $\mathbf Q$ and prove equality of the structural and tangent rotational symmetries for every nonzero coupling pair. The associated acoustic tensor links residual symmetry to protected transverse degeneracies. For a small uniaxial state, the leading squared-speed gap is $|γ_2s|\sin^2\vartheta/(2ρ_0)$; coupling selectivity and residual protection can suppress this term. We also derive a directional acoustic-trace reconstruction and state calibrated and nuisance-aware inverse conditions. Dimensionless structural-field, acoustic, transient finite-element, and synthetic inverse examples illustrate the predictions. The framework connects state evolution, mechanically expressed symmetry, and conditional observability without assigning independent dynamics to a group label.

physics.app-ph↗

arXiv2026

Context as Infrastructure: Knowledge Compilation for Retrieval in Agentic Systems with Harnesses

Retrieval-augmented generation is becoming agentic. A search harness writes queries, reads results, reformulates, and decides when to stop. This largely solves accuracy: a harness with only BM25/grep and MiniMax M3 answers 85% of BrowseComp-Plus and 91% on multi-hop FRAMES in our runs, without compiled or vector indexes. It exposes a context problem: an undisciplined loop repeats queries, rereads documents, and fills the window with unused passages. The missing layer is knowledge compilation: a configurable ingestion pipeline where LLMs enrich documents with context no chunk contains alone--navigation hierarchies with claim-evidence leaves, entity graphs, timelines, mind maps, ontologies, wikis, and memory. Compilation is not a better retriever: under a competent harness, structural artifacts gave no detectable accuracy gain on FRAMES, and a PageIndex-style outline degraded at scale. It is a supply layer. A navigation tree whose leaves are claim-evidence pairs--the mechanism we measure--cuts query tokens by 91% when evidence suffices and adds 1.9% when it does not, a 38% average reduction; this implies a 2% break-even sufficiency rate and a measured rate near 43%. A decision gate cuts latency from 140 s to 30 s, though substitution's token saving is bought with retrieval rounds and carries its own latency cost. Other artifacts--ontology, graph, wiki, timeline, mind map, memory--pay off in deployment (governance, integration, onboarding, temporal orientation, continuity) where accuracy benchmarks cannot see them. We describe RAGFlow's pipeline and harness (open source at https://github.com/infiniflow/ragflow), which turns the loop into five disciplined stations and keeps the context entering the model the smallest sufficient set; together they form an agent-native database: a store that hands the agent compiled, governed context instead of a ranked list of passages.

cs.IR↗

arXiv2026

Drone-PUR: From Visual Clue Discovery to Understanding and Reasoning in UAV Scenes

The reasoning capabilities of vision-language models (VLMs) can be limited by incomplete or inaccurate visual observations in the general domain. In unmanned aerial vehicle (UAV) scenes, small targets, subtle attributes, and changing viewpoints further complicate perception and understanding. Therefore, we introduce Drone-PUR, a dataset and benchmark organized around visual clue discovery, interpretation, and integration, corresponding to the three capability dimensions of perception, understanding, and reasoning. Reasoning samples are designed to require models to combine multiple visual clues to derive visually grounded conclusions. Its 11 tasks provide 175k training and 20k test samples spanning single images, image sequences, cross-modal images, and multiple viewpoints. All test samples undergo human review. Evaluating 22 general-purpose VLMs reveals persistent limitations in spatial judgment and cross-modal and cross-view reasoning. Across models, stronger perception is associated with better reasoning, while single-task fine-tuning further reveals selective bidirectional transfer between these capabilities. To examine domain adaptation, we develop UAV-PUR from Qwen3-VL through staged fine-tuning with auxiliary supervision. Its 4B and 8B variants improve overall multiple-choice accuracy by 18.33 and 17.52 percentage points over their pretrained counterparts, respectively. The dataset and project resources are available at https://drone-pur.github.io.

cs.CV↗

arXiv2026

String Amplitudes on Integer Rectangles

Hidden zeros of tree-level string amplitudes of colored scalars lie on loci specified by integer-valued dimensionless momentum contractions arranged in rectangles on the kinematic mesh. We study the same kinematic loci for open-string amplitudes with gluons, fermions and higher-spin states at arbitrary finite mass levels. Disk monodromy relations yield $\textit{zero-locus factorization}$: at regular points where the monodromy system is invertible, the amplitude is a weighted sum of physical three-point amplitudes sewn to complete on-shell $(n-1)$-point amplitudes. Hidden zeros follow when the external wavefunctions annihilate all contributing three-point couplings. In the low-energy limit, the massless zero-integer branch recovers the corresponding field-theory factorization. For closed strings, including gravitons, persistent cubic zeros in both chiral sectors remove the disk poles, allowing the Kawai--Lewellen--Tye momentum-kernel zero to imply amplitude zeros. The operator product expansions of external vertices give a common worldsheet origin for these conditions across spins and mass levels.

hep-th↗

arXiv2026

Orbital-coherence memory in electrical transport and its self-consistent Maxwell response

An internal electronic coherence changes electrical transport only when it feeds back on the physical current. We develop a reduced description of this feedback and distinguish its exact consequences from the assumptions required for a microscopic realization. An opposite-parity two-orbital Hamiltonian identifies a current-coupled orbital force. Under an explicitly stated moment closure and collision ansatz, it yields a passive three-state response with an exponentially damped oscillatory memory. The resulting conductivity redistributes a fixed optical weight and predicts a nonmonotonic dependence of dc conduction on coherence damping at fixed remaining parameters. The orbital reduction constrains the coupling and internal frequency through a shared bare splitting, providing a test beyond unconstrained spectral fitting. A charge-conserving spatial extension couples to Maxwell equations through a positive energy balance and separates the constitutive response from longitudinal collective and transverse electromagnetic modes. We formulate complex-conductivity and transmission tests, including the limits of parameter identification over a finite bandwidth. The model supplies a conditional connection between orbital dynamics, transport memory, and electromagnetic observables; conductivity alone does not establish a uniquely quantum mechanism, and a material-specific collision calculation and closure-error assessment remain necessary.

physics.app-ph↗

arXiv2026

Discovering Hidden Predictive States in Kinetic Transport

How do autonomous macroscopic transport laws emerge from kinetic dynamics, and which hidden degrees of freedom must survive coarse-graining? We address these questions by formulating kinetic state reduction as a problem of predictive sufficiency rather than conventional moment truncation. Starting from Boltzmann transport dynamics, we establish a quantitative connection between unresolved microscopic information, irreducible prediction error, and finite-horizon observability. Microscopic preparations with identical macroscopic observables but distinct subsequent evolution provide direct witnesses of incomplete dynamical closure. We develop an entropy-constrained learned state-reduction framework, demonstrated here through linear projections, that identifies the additional predictive states without prescribing their physical identities and determines their necessity through independent response tests, state deletion, and dynamical consistency. An exactly solvable three-velocity model demonstrates that a hidden higher-order moment can influence future energy transport despite identical initial energy and heat flux. Numerical experiments recover this missing direction and reveal how the required state dimension depends on observation scale, prediction horizon, preparation ensemble, and accuracy. In a synthetic 32-channel Boltzmann system, the learned representation identifies predictive combinations spanning multiple kinetic branches and relaxation groups that are not captured by low-order global moments. We further distinguish observable prediction accuracy from autonomous state closure and quantify the information retained through unresolved memory. These results establish an operational framework for testing the emergence of effective transport laws and identifying the microscopic information required for their predictive validity.

physics.app-ph↗

arXiv2026

CASI: Credit Assignment via Set Intervention for Agentic Retrieval

Large language models increasingly generate multiple retrieval actions jointly, while downstream systems often provide only one reward for the resulting set. Broadcasting this reward provides a valid score-function gradient estimator for the expected set-level objective, but does not determine a coalition-dependent within-set attribution. We introduce CASI (Credit Assignment via Set Intervention for Agentic Retrieval), which constructs the missing counterfactual information through snapshot-grounded set intervention: generate and retrieve once, hold replayable snapshot components fixed, and evaluate alternative subsets under common conditions. Shapley aggregation converts these marginals into within-set credit, which bounded zero-sum advantage shaping incorporates into policy optimization. For m >= 4, singleton, leave-one-out, and full-set endpoints do not generally determine the Shapley allocation; under additive set values, extra coalition evaluations are redundant. In a five-action sponsored-search setting, exhaustive auction replay over 1,200 requests reveals substantial coalition dependence. Under the training-time evaluator, CASI achieves the highest median cosine similarity to a replay-defined Shapley reference among evaluated credit rules; under a sensitivity backend, CASI does not outperform Singleton. Across five independent training runs, CASI-trained policies achieve higher held-out auction-replay set value than Singleton- or Broadcast-trained policies. In a randomized seven-day production A/B test on 10% of traffic, the CASI-trained policy yields an observed 3.51% lift in the primary advertiser-value metric over production control.

cs.AI↗

arXiv2026

Destination-Conditioned Adaptation with Semantic-Geometric Alignment for Generative Inbetweening

Generative inbetweening aims to synthesize temporally coherent intermediate frames between a start frame and a target frame. Although pretrained image-to-video diffusion models provide strong priors for large-motion inbetweening, they are not explicitly designed to reach a specified target frame. They may therefore generate realistic individual frames while drifting from the target or producing unstable motion under large displacements and non-rigid dynamics. We propose DSG-GI, a framework that adapts pretrained latent video diffusion models through \textbf{D}estination-Conditioned Adaptation and \textbf{S}emantic-\textbf{G}eometric Alignment. Destination-Conditioned Adaptation treats the target frame as the destination of the sequence and provides temporally progressive guidance along the generated trajectory. Semantic-Geometric Alignment aligns the token-wise spatio-temporal relational geometry of backbone features with that of a frozen video foundation encoder and applies a flow-matching-aware latent flow constraint to samples near the data endpoint of the probability path. Together, these components strengthen destination guidance and preserve coherent semantic-geometric motion structure. Extensive experiments on standard and large-motion inbetweening benchmarks demonstrate improvements in temporal stability, target-frame faithfulness, and perceptual quality. Our code will be made publicly available.

cs.CV↗

arXiv2026

SearchWorld: Spatial Value-Grounded Imagination for UAV Object Search via World Models

Autonomous unmanned aerial vehicle (UAV) object search involves a closed loop of perception, decision-making, and action under partial observability. Urban environments pose several challenges: large search areas and narrow egocentric views limit coverage, dense 3D geometry constrains safe motion, and open-world instructions require identifying a specific target among distractors. Many existing methods mitigate partial observability through explicit maps or memory representations, yet remain largely reactive, reasoning over past observations without explicitly predicting future states. World models enable prospective reasoning through imagined rollouts. However, image-generating world models can incur high inference latency, while spatially grounded planning remains challenging for latent world models. We propose SearchWorld, a recurrent state-space world model that connects explicit spatial memory with value-guided imagination. The model maintains BEV exploration and obstacle memory and decodes a task-aware spatial value layer to guide search. A cognition-action network uses this learned spatial value prior to improve the policy through imagined rollouts, without training a separate scalar critic. Training progresses from world-model learning to expert imitation and imagination-based exploration refinement. On UAV-ON, SearchWorld improves the success rate to 23.8% (19.5% for the strongest published agent) and raises oracle success to 35.5%, while remaining robust on unseen scenes (19.9% success rate). By grounding imagination in explicit spatial representations, SearchWorld enables UAV agents to plan prospectively rather than react.

cs.AI↗

arXiv2026

Topology-Adaptive Hyperbolic Graph Attention Networks Guided by the Hyperbolic Sombor Index

Hyperbolic geometry has emerged as a principled space for representing hierarchical graphs. However, existing hyperbolic graph neural networks typically rely on shared curvature configurations and feature-driven attention, failing to explicitly exploit local hierarchical topological patterns. To bridge this gap, we introduce the Hyperbolic Sombor Index (HSO) as a lightweight structural prior for capturing hierarchy-indicative degree stratification. Building on this, we propose \textbf{HSO-GAT}, a topology-adaptive hyperbolic graph attention network that unifies geometric adaptation and message propagation. Specifically, it comprises two complementary modules: HSO-Guided Local Curvature Adaptation, which performs adaptive node-wise geometric scaling from aggregated node-level HSO signals, and HSO-Gated Hyperbolic Graph Attention, which enables structure-aware message passing through feature-conditioned gating. Theoretically, we establish the monotonic sensitivity of edge-level HSO to degree imbalance and analyze the validity and radial scaling properties of node-adaptive hyperbolic mappings. Extensive experiments on eight benchmark datasets demonstrate that HSO-GAT consistently achieves state-of-the-art performance in both node classification and link prediction tasks.

cs.LG↗

arXiv2026

SAGE: Source-Anchored Guidance via Frequency Equalization for Hierarchical RGB-T Alignment and Fusion

Spatial misregistration and cross-modal discrepancies often cause ghosting, structural blurring, and content imbalance in RGB-T fusion. Existing methods typically decouple appearance adaptation, geometric alignment, and information fusion, limiting dependency propagation across stages. We propose Source-Anchored Guidance via Frequency Equalization for Hierarchical RGB-T Alignment and Fusion (SAGE), a unified framework integrating frequency equalization, hierarchical alignment, and subband fusion. SAGE employs invertible joint encoding and source-specific low-frequency modulation to derive structural and gain guidance while preserving source information. Hierarchical frequency collaborative alignment estimates global affine geometry from low-frequency approximations and transfers geometric and contextual cues to high-frequency correlation reasoning for reliability-aware residual refinement. Guided subband fusion jointly aggregates the aligned frequency coefficients under propagated source and alignment guidance, coordinates complementary low- and high-frequency information, and reconstructs the fused image through the inverse wavelet transform. Extensive experiments on RGB-T datasets with real-world and synthetic misalignments demonstrate consistently competitive performance in alignment and fusion, validating the effectiveness of source-anchored guidance for weakly registered RGB-T images.

cs.CV↗

arXiv2026

Socialized UAV Cross-Task Learning: Towards Cross-Granularity Collaboration through Hierarchical Interaction

Joint learning across heterogeneous tasks is often treated as task coupling through feature sharing, distillation, or auxiliary supervision. However, in cross-task learning, mismatched representational and supervisory granularities make such coupling prone to interference, teacher bias, or unidirectional collapse. We argue that cross-granularity learning is fundamentally a problem of hierarchical interaction regulation rather than simple task coupling. This issue is particularly evident in UAV perception, where visual shifts and detection--segmentation objectives naturally form coarse- and fine-grained knowledge sources. To systematically study this problem, we introduce CrossUAV, a UAV benchmark for joint object detection and instance segmentation that provides a unified evaluation platform for cross-granularity task collaboration. To address these challenges, we propose Cross-Granularity Socialized Collaboration (CGSC), a progressive and adaptive framework that regulates when, where, and how tasks exchange information across network hierarchies. CGSC progressively activates cross-task interactions and adaptively adjusts the strength according to task contribution, suppressing harmful interference while exploiting complementary coarse- and fine-grained structures. Extensive experiments demonstrate consistent improvements on both tasks, validating hierarchical dynamic interaction as an effective mechanism for cross-granularity collaboration.

cs.CV↗

arXiv2026

Can Deep Learning Achieve Cross-Physics Mapping?

Can deep learning translate physical fields governed by fundamentally different equations? We address this question by introducing Cross-Physics Mapping (CPM), an operator-learning framework for mappings between heterogeneous physical domains. We formulate sufficient conditions for such mappings through compatible latent representations and propose a dimensionless scaling principle that aligns the characteristic evolution scales of the source and target systems without assuming their dynamical equivalence. As a representative test, paired diffusion and wave fields are generated independently from their respective parabolic and hyperbolic equations while sharing the same latent geometry, material heterogeneity, excitation, and dimensionless scale. Seven architectures-ResUNet, DeepONet, Fourier, latent, wavelet, U-shaped, and Galerkin neural operators-are evaluated for both diffusion-to-wave and wave-to-diffusion mappings. The results reveal a strong directional asymmetry. Diffusion-to-wave reconstruction is more challenging because it requires recovering wavefront, phase, and time-of-flight information attenuated by diffusion; U-NO performs best in this direction, achieving a relative $\ell_2$ error of $0.307$ and an $R^2$ of $0.905$. Wave-to-diffusion mapping is considerably more stable, with GNO attaining a relative $\ell_2$ error of $0.154$ and an $R^2$ of $0.935$. Neural operators generally outperform the conventional convolutional baseline, highlighting the nonlocal nature of cross-physics transformations. These findings demonstrate that deep learning can establish useful mappings between distinct physical modalities on a shared latent manifold, while the achievable accuracy remains fundamentally constrained by the direction-dependent information content of the governing physics.

cs.LG↗

arXiv2026

A2DINOv3: Rethinking Multi-Modal Object Detection via Socialized Collaboration

Multi-modal object detection is essential for robust scene understanding in challenging conditions, including low-light and adverse environments. Recent vision foundation models (e.g., DINOv3) have exhibited strong representation capabilities, yet adapting them to multi-modal scenarios remains challenging. Existing dense cross-modal fusion strategies often force heterogeneous modalities to interact indiscriminately, which may introduce redundant information and disrupt the valuable pre-trained representations. To address this issue, we revisit multi-modal fusion from the perspective of socialized learning and propose adapter to DINOv3 (A2DINOv3), a multi-expert collaboration framework with a Socialized Collaboration Protocol (SCP). Specifically, RGB and infrared branches are modeled as heterogeneous experts that independently preserve their specialized knowledge while exchanging complementary information through selective and constrained interactions. This design mitigates harmful cross-modal interference and prevents degradation of pre-trained priors during adaptation. Furthermore, a zero-initialization strategy is introduced to gradually activate cross-modal collaboration, enabling a smooth transition from modality-specific learning to cooperative representation learning. Extensive experiments on four multi-modal benchmarks, including aerial detection (GAIIC), autonomous driving (FLIR), low-light surveillance (LLVIP), and diverse real-world scenarios (M3FD), demonstrate that A2DINOv3 consistently achieves state-of-the-art performance in multi-modal object detection.

cs.CV↗

arXiv2026

Spectral Geometry and Dispersion-Constrained Projection of Diffusive Fields

Diffusive fields obey operator-imposed relations between spatial structure and temporal decay, yet conventional spectral filtering selects components primarily according to frequency or wavenumber magnitude. Here we show that the diffusion operator defines a spectral geometry in the joint space of spatial wavenumber and modal decay rate, where physically admissible modes occupy the manifold $η=α|\mathbf{k}|^2$. This geometry separates spectral scale from physical consistency: high-wavenumber modes can remain diffusion-consistent, whereas lower-wavenumber modes can violate the governing dynamics. We exploit this distinction by introducing an operator residual and a finite-width soft projection that selects spectral components according to their distance from the diffusion manifold rather than their spectral magnitude. Numerical studies demonstrate robust recovery under noise, diffusivity mismatch, and finite acquisition, and reveal a consistency--retention tradeoff governed by the manifold width. Photothermal experiments further confirm that the projection suppresses off-manifold spectral content while retaining the dominant thermal response. These results establish operator consistency as a spectral-selection principle for diffusive fields and provide a geometric framework for physics-informed processing of dissipative systems.

physics.app-ph↗

arXiv2026

Fermionic hidden zeros

We uncover and prove a general class of hidden zeros in tree-level amplitudes with massless fermions. For arbitrary even numbers and arbitrary arrangements of massless real adjoint fermions and gluons, color-ordered gauge-theory amplitudes in $D=4,6,10$ vanish on general rectangular kinematic loci supplemented by species-dependent bridge conditions. The fermion--fermion bridge is governed simply by the vector current $χ_iγ^μχ_j$, while mixed and gluonic bridges are fixed by the corresponding gauge-covariant contractions. A worldsheet analysis proves the result at arbitrary multiplicity and traces its universality to the local fusion channels $ff\to g$, $fg/gf\to f$, and $gg\to g$, with no new primitive bridge structures appearing at higher fermion multiplicity. The same mechanism extends to ten-dimensional amplitudes with gravitons and arbitrary even numbers of gravitinos in factorized gamma-traceless polarizations, and to single-trace Einstein-Yang-Mills amplitudes with adjoint gluinos and gravitons. These results expose a finite local structure underlying fermionic hidden zeros across gauge and gravitational amplitudes.

hep-th↗

arXiv2026

UniDiffFusion: A Unified Diffusion Framework for Multi-Task and Degradation-Robust Image Fusion

General image fusion aims to integrate complementary information from multiple source images, but existing methods often rely on task-specific models and struggle to maintain robust performance under diverse degradation conditions. In this paper, we propose UniDiffFusion, a unified diffusion framework for multi-task and degradation-robust image fusion. UniDiffFusion leverages the strong generative prior of a pretrained diffusion model to establish a shared fusion backbone across heterogeneous fusion tasks, while introducing task- and degradation-aware conditional adaptation to accommodate their distinct information-selection requirements. Specifically, we employ task prompt modulation to progressively adapt the shared diffusion representations to different fusion objectives, and develop a degradation prompt router to dynamically retrieve degradation-aware priors and restore corrupted source features before fusion. Furthermore, an application prompt bank is introduced to incorporate task-oriented semantic guidance for downstream applications, such as object detection and semantic segmentation, without altering the shared fusion and restoration pathways. The proposed framework is trained in a progressive manner to decouple fusion learning, degradation-aware restoration, and application-specific adaptation, thereby reducing interference among heterogeneous objectives. Extensive experiments on visible-infrared, multi-exposure, and multi-focus image fusion demonstrate that UniDiffFusion achieves superior fusion quality and robustness under both clean and degraded conditions. Moreover, UniDiffFusion consistently improves downstream detection and semantic segmentation performance, demonstrating its effectiveness as a unified diffusion framework for both perceptual fusion and task-oriented vision.

cs.CV↗

arXiv2026

A-PAIR: A Benchmark and Identity-Consistent Grounding Framework for Air-Ground Cross-View Referring Person Detection

Air-ground cross-view referring person detection is a necessary component in the language-to-perception-to-control chain of collective embodied intelligence, grounding a language command into the same physical target before ground and aerial agents can coordinate downstream actions. Existing referring expression comprehension and open-vocabulary grounding methods do not jointly account for cross-view identity consistency, making them insufficient for Air-Ground Cross-View Referring Person Detection (AGCV-RPD), which involves similar pedestrian distractors, weak aerial appearance cues, and cross-view identity consistency. To study this problem, we introduce Air-Ground Paired Identity-Aware Referring (A-PAIR), the first comprehensive AGCV-RPD benchmark, containing 22,137 cross-view referring samples. To construct A-PAIR efficiently, we propose Factorized Annotation and Referential Alignment (FARA), a semi-automatic annotation framework that generates factorized referring descriptions and identity-consistency supervision at reduced cost. We propose Identity-Consistent Referring Grounding (ICRG), a framework that combines factorized referential grounding, candidate-completeness supervision, and cross-view consistency calibration for joint air-ground pair selection. ICRG improves ground, aerial, and pair-level detection over strong baselines, increasing pair F1 from 16.65% to 22.28%. These results show that AGCV-RPD requires paired detection and identity-consistent reasoning.

cs.CV↗