arXiv subjects

Yifeng Zhang

Publications and source records attributed to Yifeng Zhang.

At least 19 recordsLinked to original sources

arXiv2026

BEVLight: Generalizable Traffic Signal Control from UAV Bird's-Eye Views

Learning-based traffic signal control (TSC) has achieved promising performance, but most methods rely on structured traffic states that must be estimated from sensor observations in real deployments. Visual TSC learns signal decisions directly from images, yet existing methods depend on fixed roadside cameras and remain tied to specific intersection configurations and visual domains. UAVs offer on-demand bird's-eye-view (BEV) observations that jointly capture intersection topology and traffic dynamics, making them a promising sensing platform for visual TSC. We present BEVLight, a generalizable TSC framework that directly maps BEV observations to signal actions. BEVLight grounds visual features onto individual lanes and composes them along the junction topology into movement and phase representations, allowing a shared policy to score candidate phases across heterogeneous intersection layouts and signal plans. To train the policy, we develop a web-based virtual teleoperation platform and collect over 5K human control decisions together with their underlying traffic states. Simulator-provided traffic states first supervise lane- and movement-level representations, after which human phase selections and rankings supervise the phase scorer. Re-rendering the recorded trajectories under diverse appearances preserves both forms of supervision and encourages robustness to visual changes. Closed-loop experiments demonstrate zero-shot generalization across traffic demands, signal plans, and intersections: BEVLight reduces average waiting time by 7--9\% relative to perception-to-control visual baselines and matches or outperforms structured-state RL controllers that receive ground-truth traffic states, achieving the lowest waiting time in 12 of 16 scenarios. An offline study further shows that BEVLight transfers to real UAV observations without target-domain adaptation.

eess.SY↗

arXiv2026

Correlation-Aware Exploration for Long-Trace Reasoning

Reasoning models trained with Group Relative Policy Optimization (GRPO) now produce chains of thought thousands of tokens long, and recent studies reveal that such training is prone to policy collapse. Entropy regularization is the standard remedy, but it works by pushing probability mass away from the few tokens that carry it and toward the vast low-probability tail of the vocabulary, and it applies this force once for every token of the reasoning trace. This puts the bonus in direct conflict with the entropy decay that convergence requires, and makes its coefficient difficult to set in the long-trace regime. To resolve this tension, we propose Info-GRPO, an information-theoretic framework that cultivates correlation between the policy and a latent prior. Rather than raising entropy everywhere, Info-GRPO builds on the mutual information between a latent variable and the trajectory it induces: augmenting prompts with latent seeds during training lets the model explore diverse policies correlated with the prior, while the seed-conditioned entropy is guided toward convergence. Diversity is therefore carried across whole reasoning traces rather than by per-token randomness, and exploration no longer fights entropy reduction. Experiments show that Info-GRPO outperforms vanilla GRPO and entropy-regularized GRPO across diverse reasoning benchmarks. On AIME24, it improves $Avg@8$ over GRPO by 3.75%, 1.66%, 4.16%, and 4.17% on four backbones, and it reaches the same majority-vote accuracy with $8\times$ fewer samples. Code and models will be publicly released.

cs.LG↗

arXiv2026

PhantomTac: Hijacking Robotic Tactile Systems with Deploy-and-forget Tags

Tactile perception is crucial for precise robotic manipulation, providing an important foundation for dexterous, closed-loop force control. Among various tactile sensing techniques, magnetic tactile sensors stand out thanks to their fine-grained, three-dimensional force perception capability and strong robustness against visual occlusion. In this study, we discover a physical loophole called PhantomTac in magnetic tactile sensors. The attacker can alter the magnetic field, thus spoofing the force interpreted by magnetic tactile sensors. The attacker can stealthily deploy such an attack by simply attaching a thin (the thickness is around 0.5 mm), battery-free magnetic tag to the surface of an object. Hence, the attack tag can be easily concealed by a barcode sticker. Such an attack scheme is "deploy-and-forget" -- the attacker can deploy the tag and leave. When a robotic hand is picking up an object, it may severely damage the object. Specifically, spoofing the normal force can crush the object, whereas spoofing the shear force can lead to a loose grip, thus dropping the object. Compared with directly sabotaging the robotic hand, PhantomTac is much more concealed as the attacker does not need access to the device. We performed extensive reverse-engineering efforts to characterize the sensing mechanism of commercial-off-the-shelf tactile sensors and proposed a practical tag design. End-to-end evaluations conducted on real robots and robotic arms demonstrate that our attack can bypass existing anti-interference mechanisms and effectively hijack their closed-loop control processes, achieving an attack success rate exceeding 95% when the grasping-position offset is bounded within 8 mm. Finally, based on the limitations of existing anti-magnetic interference mechanisms, we propose a multimodal-consistency defense mechanism to mitigate such physical-layer security threats.

cs.CR↗

arXiv2026

From Impact to Insight: Dynamics-Aware Proprioceptive Terrain Sensing on Granular Media

Robots that traverse natural terrain must interpret contact forces generated under highly dynamic conditions. However, most terrain characterization approaches rely on quasi-static assumptions that neglect velocity- and acceleration-dependent effects arising during impact and rapid stance transitions. In this work, we investigate granular terrain interaction during high-speed hopping and develop a physics-based framework for dynamic terrain characterization using proprioceptive sensing alone. Through controlled hopping experiments with systematically varied impact speed and leg compliance, our measurements reveal that quasi-static based assumptions lead to large discrepancies in granular terrain property estimation during high-speed hopping, particularly upon touchdown and controller-induced stiffness transitions. Velocity-dependent drag alone cannot explain these discrepancies. Instead, acceleration-dependent added-mass effects - associated with grain entrainment beneath the foot - dominate transient force responses. We integrate this force decomposition with a momentum-observer-based estimator that compensates for rigid-body inertia and gravity, and introduce an acceleration-aware weighted regression to account for increased force variance during high-acceleration events. Together, these methods enable consistent recovery of granular resistance across locomotion conditions, closely matching linear-actuator ground truth. Our results demonstrate that accurate terrain inference during high-speed locomotion requires explicit treatment of acceleration-dependent granular effects, and provide a foundation for robots to characterize complex deformable terrain during dynamic exploration of terrestrial and planetary environments.

cs.RO↗

arXiv2026

MiniWAM: Learning Compact Future Targets for Efficient World-Action Modeling

World modeling has emerged as an effective co-training objective for robot policies, giving rise to World Action Models (WAMs) that jointly predict actions and future states. However, most WAMs predict future states in the native representation space of pretrained visual backbones, resulting in high-dimensional targets with substantial training cost. We introduce MiniWAM, which instead predicts compact future representations learned from privileged current-future transitions. To construct these targets, we propose Predictive Representations via Inverse Spatiotemporal Modeling (PRISM), which combines inverse-dynamics supervision with feature reconstruction to emphasize control-relevant transition information while preserving useful future-state information. With the learned PRISM encoder frozen, MiniWAM is trained to jointly predict the resulting targets and robot actions from current observations. With 65$\times$ fewer native future feature tokens, MiniWAM consistently outperforms native future-feature prediction with both DINOv3 and WAN2.1 VAE features, while achieving up to an 8$\times$ speedup in world-action training. At 0.25B parameters, MiniWAM is already competitive with substantially larger WAMs on LIBERO, LIBERO-Plus, and RoboTwin 2.0 simulation benchmarks. Representation analyses further show that PRISM contributes behavioral structure beyond feature reconstruction alone. These results demonstrate that effective world-action modeling does not require predicting native visual futures, and that compact predictive representations provide a strong and substantially more efficient target for policy learning. The project page is available at: https://j1dan.github.io/MiniWAM.

cs.RO↗

arXiv2026

TAGA: Terrain-aware Active Gaze Learning for Generalizable Agile Humanoid Locomotion

Agile humanoid locomotion across diverse challenging terrain demands both wide perceptual coverage and precise local geometry understanding. Motivated by the way humans selectively look at relevant terrain during locomotion, we introduce TAGA, a Terrain-aware Active Gaze learning framework for Attention-based humanoid control. By fusing vision, proprioception, and motion commands, our framework guides the model to learn anticipatory cues and actively attend to specific areas of the height scan, selectively using these informative regions for the downstream network. This adaptively increases the information density of observations under tight onboard computational constraints, thus enabling fine-grained perceptive locomotion over larger-scale terrains. We find that such gaze behaviors can naturally emerge through reinforcement learning alone, without requiring additional supervision or explicit guidance, significantly improving training efficiency. As a result, the trained policy demonstrates robust and generalizable locomotion in simulation and on hardware, including reliable terrain-aware foothold selection, elevated-platform traversal, competitive sparse-foothold traversal, and the largest reported real-world gap traversal distance of 1.2m among perceptive humanoid locomotion systems, while maintaining stability under severe perceptual disturbances and environmental interference.

cs.RO↗

arXiv2026

Orchestrating Wheeled Mobile Robots Online without Conflicts

Formation control of wheeled mobile robots (WMRs) has been widely studied due to its broad applications. However, existing studies primarily focus on tracking predefined formations, with limited adaptability to varying environments. To address this limitation, we propose CFOO, a Conflict-Free Online Orchestration framework for continuous formation navigation of WMRs that integrates centralized formation generation with distributed formation maintenance. The upper layer monitors the environment in real time and generates adaptive formations as needed, using the proposed TCF-R2T (Trajectory-Conflict-Free Robot-to-Target assignment) algorithm to compute optimal assignments in polynomial time for timely, conflict-free formation transitions. At the lower layer, each WMR continuously employs JSTP (Joint Spatio-Temporal trajectory Planning) to maintain the given formation by jointly optimizing spatial positions and segment durations, improving coordination over spatial-only optimization with fixed durations. By coordinating these two layers, CFOO orchestrates WMRs to adapt their formations to varying environments and thus enable continuous navigation. Moreover, we design a Time-Varying Formation Reference (TVFR) mechanism to facilitate smooth transitions between successive desired formations. Both simulation and real-world experiments validate the effectiveness and adaptability of CFOO. Experimental videos are available on our project website: https://github.com/dongjh20/CFOO.

cs.RO↗

arXiv2026

Quasiparabolic Gelfand models for finite irreducible Coxeter groups and their canonical Hecke structures

Quasiparabolic sets extend parabolic coset spaces while retaining a length filtration and natural Hecke algebra deformations. Gelfand models built from induced linear characters ask when such spaces can account for every irreducible representation exactly once. We classify, up to equality of the individual induced characters, all quasiparabolic Gelfand models for finite irreducible Coxeter groups with inducing characters restricted from their ambient parabolic subgroups, including the additional models of type $D_{4r+2}$ and $B_3$. We derive the associated finite Gelfand-pair and commutant consequences, construct canonical Hecke structures for the additional classical models.

math.RT↗

arXiv2026

SMaRT-Tug: Structured Multi-Agent Reinforcement Learning for Physics-Based Tugboat-Barge Collaborative Manipulation

Autonomous tugboating is central for automating maritime operations such as port logistics and vessel maneuvering, where multiple tugboats must cooperatively transport/manipulate a larger vessel. Collaborative pushing in this setting is challenging due to coupled hydrodynamics, low resistance, strong environmental disturbances, underactuated barge dynamics, and contact-rich interactions. Conventional control methods often rely on simplified models and fixed configurations, which limit their adaptability, while learning-based approaches are constrained by the lack of scalable and physically realistic training environments. We address these challenges by introducing a physics-based, GPU-accelerated simulation and learning framework for collaborative tugboat manipulation. Our simulator incorporates a customized buoyancy model, wave modeling, and hydrodynamic resistance, and supports large-scale multi-agent training under marine dynamics. In this simulator, we train a decentralized MAPPO (Multi-Agent PPO) policy augmented with a structured control prior (SCP) to improve training stability and maintain feasible pushing configurations. We evaluate our learned policy on straight-line transit, turning, and deceleration tasks, where we show that our decentralized framework yields more reliable and accurate maneuvering performance compared to a PID-based controller and a centralized PPO baseline. We further demonstrate zero-shot generalization to more challenging sea states and advanced maneuvers, as well as zero-shot scalability to larger teams of three and four tugboats despite training with only two agents.

cs.RO↗

arXiv2026

Latency-Aware Orchestration for Multi-Agent LLM Workflows on Heterogeneous GPUs

Concurrent multi-agent workflows expose future dependencies and serving-state requirements while running on heterogeneous GPU pools with time-varying load, model residency, and resource availability. The logical workflow defines the required computation, whereas its physical scheduling units, model-lifecycle actions, resource ordering, and placement must be selected according to the observed pool state. We present a prediction-guided runtime that uses workflow forecasts to construct and optimize a physical execution graph. Predictor estimates device-specific activation latency, peak memory, and model-loading cost, then propagates these predictions through workflow dependencies to forecast activation readiness and future model demand. Constructor builds semantics-preserving fusion and model-lifecycle alternatives, while Scheduler jointly optimizes their selection, placement, and execution order based on the live pool state. Across a workload spanning three workflow scenarios on a heterogeneous GPU pool, our system reduces end-to-end makespan and overall p95 completion latency under burst arrivals by up to 36.8% and 25.9%, respectively, over state-of-the-art workflow schedulers. It also saves up to 24.63 GPU-s per completed session.

cs.DC↗

arXiv2026

Cell Classification of Gelfand $S_n$-Graphs

Kazhdan--Lusztig introduced $W$-graphs to encode Hecke algebra representations on canonical bases, whose strongly connected components are the cells of the graph. A canonical basis admits additional structure under restriction to Levi quantum subalgebras: the connected components of the corresponding Levi crystals have dominant highest weights, called Levi tops, which organize the basis through dominance filtrations. In this paper, we establish a general multiplicity-free Levi-top criterion for canonical-basis Hecke graphs. When the Levi restriction is multiplicity-free, the relevant dominance filtrations are compatible with the Hecke action, and each Levi-top fiber is connected by bidirected edges, the cells, molecules, and Levi-top fibers coincide. Applying this criterion to the two type-$A$ Gelfand $S_n$-graphs arising from the row and column Beissinger correspondences, we prove that every cell is exactly one molecule.

math.CO↗

arXiv2026

Molecules of an affine FPF $W$-graph and a labelled row-Beissinger reconstruction

Kazhdan--Lusztig $W$-graphs encode the cell structure of Hecke algebras, while their bidirected connected components are called molecules. In finite type~$A$, the Robinson--Schensted correspondence describes cells and molecules, and Beissinger's row insertion constructs the common tableau associated with an involution. In affine type~$A$, the affine matrix-ball construction (AMBC) assigns an affine permutation a pair of tabloids together with a dominant weight, and Marberg introduced affine fixed-point-free (FPF) $W$-graphs indexed by affine FPF involutions. We classify and enumerate the molecules of Marberg's $\m$-type affine FPF $W$-graph $Γ_n^{\m}$ and show that molecules with the same AMBC shape have isomorphic underlying simple bidirected graphs. We also prove that labelled complete-cycle row-Beissinger truncations recover the full AMBC datum of an affine FPF involution.

math.CO↗

arXiv2026

Molecules of Type-$B$ and Odd Type-$D$ Gelfand $\mathbf m$- and $\mathbf n$-Graphs

Gelfand $W$-graphs provide multiplicity-free canonical models for classical Weyl groups, and their molecules record the connectivity generated by bidirected canonical-basis edges. Restricted Beissinger insertion and dual equivalence give a natural tableau language for these edges, while the wall and fork generators require additional local criteria. For both type-$B$ Gelfand $\mathbf m$- and $\mathbf n$-graphs, we prove that the molecules are exactly the connected components of explicit occurrence--wall graphs constructed from split modified Beissinger insertion. In odd type $D$, we likewise classify the $\mathbf m$-molecules by a completion-filtered occurrence--fork graph and obtain the $\mathbf n$-classification through the parity-sensitive type-$D$ duality.

math.RT↗

arXiv2026

CGFM-Nav: Cognitive Graph-Field Memory for Semantic-Guided Lifelong Multimodal Embodied Navigation

Vision-and-Language Navigation (VLN) requires agents to reason over accumulated observations while continuously exploring unseen regions. However, existing environment representations often struggle to jointly support explicit semantic memory and continuous exploration guidance. To address this challenge, we propose Cognitive Graph-Field Memory (CGFM), a persistent multimodal scene representation that couples explicit relational memory with continuous spatial intuition. CGFM organizes objects, spatial relations, and visual observations into a multimodal scene graph, enabling target retrieval and long-horizon reasoning across navigation tasks. When no reliable target match is identified, graph-based evidence is projected into a goal-conditioned semantic-frontier field to guide exploration toward semantically promising frontiers and regions. Building upon CGFM, we introduce CGFM-Nav, a foundation-model-based framework for lifelong multimodal navigation that integrates task-relevant subgraph selection, VLM reasoning, and verification feedback into a closed decision loop. Preliminary experiments on GOAT-Bench show that, under the same Qwen3-VL-8B backbone, CGFM-Nav improves the overall success rate from 53.2% to 63.0% and SPL from 30.0% to 39.6%, demonstrating the effectiveness of combining explicit semantic memory with semantic-guided exploration.

cs.RO↗

arXiv2026

Dual Weight and Monodromy of Dual Affine RS Correspondence

The dual affine Robinson--Schensted correspondence and the affine matrix-ball construction give two related parametrizations of extended affine permutations. From the stable-window data of the dual correspondence, we introduce a dual weight $β$ and prove that it is consistent with the original pair $(λ,N_0)$. We prove that $β$ and the AMBC weight $ρ$ have identical monodromy along affine Knuth paths. We further give an explicit relation between $β$ and $ρ$, showing that their difference depends only on the associated tabloids.

math.CO↗

arXiv2026

Recursive structures of molecules and cells in Gelfand $S_n$-graphs

$W$-graphs, representing the multiplication action of the standard basis on the canonical basis in the Iwahori-Hecke algebra are introduced by Kazhdan and Lusztig. Marberg defined a generalized $W$-graph, the Gelfand W-graph, corresponding to the Hecke algebra modules instead of Hecke algebras. To classify the molecules and cells of the Gelfand $S_n$-graphs, in this paper, we introduce a recursive structure of $S_n$ and then discuss the action of the recursive structure on the molecules. Using this structure, we prove that the distinguished nonnesting molecule is a cell.

math.CO↗

arXiv2026

Lusztig $\mathbf{a}$-functions for quasiparabolic sets

Kazhdan and Lusztig introduce the $W$-graphs to describe the cells and molecules corresponding to the Coxeter groups. Building on this foundation, Lusztig defines the a-funtion to classify the cells, as well as the molecules. Marberg then generalizes Kazhdan and Lusztig's $W$-graphs, using fixed-point-free involutions as their indices. The molecules of the two new $S_n$-graphs are then classified via two correspondence similar to RSK correspondence by Marberg and me. In this paper, we define an analogue of the Lusztig a-function and finish the classification of cells by proving that every molecule in the $S_n$-graphs is indeed a cell.

math.CO↗

arXiv2026

JD Oxygen AI Item Center (Oxygen AIIC) V1: An Industrial-Scale LLM/VLM-Centric Solution for Item Understanding, Management, and Applications

JD$.$com, one of the world's largest e-commerce platforms, serves over 700 million active users and millions of merchants, with a catalog of tens of billions of SKUs. At this scale, high-quality, structured item knowledge underpins a better consumer experience, lower management costs, and higher operational efficiency-yet producing and serving it poses three industrial-scale challenges: fast-emerging concepts, high-quality knowledge production for massive SKUs, and diverse downstream requirements. To address these challenges, we present the JD Oxygen AI Item Center (Oxygen AIIC), an industrial-scale platform built on LLMs/VLMs for item-knowledge production and service. Oxygen AIIC is built around four core pillars: (i) ontology engineering driven by efficient human-AI collaboration, which supports the dynamic evolution and agile expansion of an ontology with millions of entries; (ii) a "Semantic Search then Discrimination"(S2D) knowledge identification architecture that, combined with throughput improvement strategies, enables scalable, extensible, and high-throughput AI Item Library production for tens of billions of SKUs; (iii) self-evolving item-understanding LLMs/VLMs that improve in a stable and controllable manner, enabling knowledge production with 94.2% precision and 82.8% recall; and (iv) a unified item tunnel that serves as the data and service hub. Oxygen AIIC now covers tens of thousands of JD categories and processes hundreds of millions of item updates per day on Huawei Ascend NPUs. It has accumulated hundreds of billions of item-knowledge assets. Deployed across core business scenarios-including search, recommendation, operations, category planning-Oxygen AIIC has delivered measurable gains at scale. Search-traffic coverage reaches 80.4%, item-information quality issues drop by 37%, the automated fill rate of core attributes during item listing exceeds 80%.

cs.AI↗