Searcharxiv⌕ Search

arXiv subjects

Yunxi Jiang

Publications and source records attributed to Yunxi Jiang.

7 recordsLinked to original sources

LEASH: Stay Safe in Agentic Robotics

Foundation models such as GPT and Claude have shown strong capabilities in spatial understanding and embodied intelligence. These advances make agentic robotics a promising path toward artificial general intelligence in robotics. However, we find that using foundation models to control robots poses serious safety risks, and comprehensive safety evaluation is still lacking. We introduce RoboRisk, a comprehensive benchmark for evaluating safety in agentic robotics. Its 92 simulated tasks and 13 physical-robot tasks fall into five categories: Task-level Refusal (T), Plan Repair (P), Execution-time Hazard (E), Injection (I), and Sequential Memory (S). Prior benchmarks cover at most two of these categories, and several evaluate only text outputs. In contrast, RoboRisk covers all five and executes every task in closed loop. We evaluate six state-of-the-art foundation models on RoboRisk: GPT-6 Astra, GPT-6.1 Sol, GPT-6 Luna, Claude Opus 5.5, Gemini 4 Argon, and Grok 4.7. These models generally show high harm rates. We also evaluate two robot-specific APIs: the Dream API and PokeBot's RoboAPI. We further propose the Lease-Enforced Agentic Safety Harness (LEASH), which offers essential safety infrastructure for agentic robotics and can be integrated into nearly any agentic architecture. LEASH leverages a decision model (Jev or OpenAI's Decisions API) to implement task, action, and commit gates, which review the task, each action, and each irreversible operation in their visual context. The planner and all three gates share a common safety policy. Experiments on RoboRisk in simulation and on physical robots show that LEASH effectively reduces harm across all six foundation models. It largely preserves their task completion performance and adds little time and token overhead.

cs.RO↗

Spin-conserving and spin-flip magnetic scattering in the altermagnetic spin-splitter effect

Critical magnetic fluctuations scatter electrons strongly near an ordering transition, but they barely degrade the charge current, because they scatter mostly through small angles. In an altermagnet the electric field also drives a spin current, and a nearly forward spin flip moves an electron to the other spin band at almost unchanged velocity. We develop a Boltzmann theory of the spin splitter ratio that keeps the full collision operator of quasistatic magnetic fluctuations and its change with the magnetic order. Fluctuations of equal total strength give the same resistivity for every fraction of spin flips, yet they can move the spin splitter ratio in opposite directions. Spin conserving fluctuations relax the spin current at about the charge relaxation rate, whereas forward flips equalize the two spin populations at equal energy at the quasiparticle rate, which in two dimensions grows linearly with the correlation length. Energy conservation on the split bands protects the part of the spin current carried by the band shift of the occupied states at every scattering strength, and the splitting blocks forward flips once the spin Fermi lines lie farther apart than the inverse correlation length. When flips dominate, a quarter to a half of the lattice band ratio survives. The transport time approximation reproduces the initial growth of the resistivity but recovers only a quarter to a half of the initial fall of the ratio, and a disorder series at fixed reduced temperature measures this fall independent of the order amplitude.

cond-mat.mes-hall↗

Spin conservation protects the altermagnetic spin-splitter current from critical scattering

Near a magnetic transition, critical fluctuations scatter electrons strongly, but the charge current survives, because long wavelength fluctuations barely turn the electron velocity. In an altermagnet the electric field also drives a spin current, the spin splitter effect, and a nearly forward scattering event that flips spin moves an electron to the other spin band at almost unchanged velocity. We show that the spin current shares the protection of the charge current only when the fluctuations conserve spin. Solving the Boltzmann equation with the full collision operator, including its change with the magnetic order, we find that forward flips equalize the two spin populations at equal energy at the quasiparticle rate, which in two dimensions grows linearly with the correlation length, whereas spin conserving fluctuations relax the spin current at a rate comparable to the charge relaxation rate. Fluctuations with the same resistivity can thus move the spin splitter ratio in opposite directions. Energy conservation protects the part of the spin current that follows the band shift of the occupied states, and the splitting blocks forward flips once the spin Fermi lines lie farther apart than the inverse correlation length. A disorder series at fixed reduced temperature measures the initial fall of the ratio with the magnetic resistivity independent of the order amplitude, and the growth of this slope toward the transition identifies spin flip critical fluctuations.

cond-mat.mes-hall↗

PanoPed: Beyond Bounding Boxes for Sim-to-Real Panoramic Pedestrian Tracking

Full-sphere panoramic cameras let fixed monitoring systems and mobile robots track people in every direction, but a planar bounding box does not fully describe where a person is on the sphere. We introduce PanoPed, a sim-to-real benchmark for pedestrian tracking on the full sphere. PanoPed-S contains 108,000 frames from fixed, quadruped-mounted, and drone-mounted cameras, with synchronized masks, depth, camera poses, and 3D pedestrian states. PanoPed-R adds 28,002 real frames from fixed cameras, 16,247 of them densely annotated. We find that an ERP rectangle cannot uniquely determine the spherical center and angular extent of the visible person, while the detector's visual query still carries information about them. Inspired by the sextant's use of angular measurements to locate objects, we propose Sextant, a plug-and-play angular localization head with only about 0.035M parameters. It reuses a frozen detector, keeps track identities unchanged, and needs no extra image encoder. Sextant gives the best result in our PanoPed-S test comparison, raising the strongest baseline, MOTIP, from 47.30 to 49.49 HOTA, with gains on all eight test sequences. Without fine-tuning on real data, the same synthetic-trained heads improve MOTIP and HAT by 0.96-1.14 HOTA on real video, and both seeds improve every real sequence. HAT+Sextant scores best among the compared systems that add no localization image encoder.

cs.CV↗

Antisymmetric breathing in altermagnetic skyrmions

A skyrmion in an altermagnet with \(d\) wave symmetry consists of two elliptical sublattice textures with perpendicular long axes. For each sublattice component $η=A,B$, we define an effective skyrmion radius $R_η=\sqrt{a_ηb_η}$, where $a_η$ and $b_η$ are the distances from the skyrmion center to its boundary along $y$ and $x$, respectively. For the anisotropic exchange parameters studied, the equilibrium radius at zero field is smaller than in a reference with parallel sublattice textures and otherwise identical parameters. A magnetic field perpendicular to the film expands one sublattice skyrmion and contracts the other, generating a radius difference $\dR=R_A-R_B$ and a net magnetic moment. The exchange modulation used to represent uniaxial strain also produces a nonzero radius difference at zero field. Because the strong and weak exchange directions are interchanged between the two sublattices, a common directional change of the exchange couplings increases one radius and decreases the other. Unlike the magnetic coupling, this mechanism vanishes when the two sublattice exchange tensors become identical. The radius difference also supports an antisymmetric breathing mode. After a short field pulse, it oscillates in quadrature with the uniform helicity, the common rotation of the wall magnetization within the film plane, at \(49.5\,\mathrm{GHz}\) for the reference parameters.

cond-mat.mes-hall↗

SO3UFormer: Learning Intrinsic Spherical Features for Rotation-Robust Panoramic Dense Prediction

Panoramic dense-prediction models, spanning semantic segmentation and depth estimation, are typically trained under a strict gravity-aligned assumption. Real-world captures, however, routinely violate it: handheld devices jitter and aerial platforms change attitude, so the camera is rarely upright. Under such 3D reorientation, standard spherical Transformers overfit global latitude cues and collapse. We introduce SO3UFormer, an architecture that learns intrinsic spherical features largely decoupled from the underlying coordinate frame, through three geometric components: (1) removing absolute latitude encoding, which breaks the dependence on the gravity axis; (2) quadrature-consistent spherical attention, which corrects for non-uniform sampling density; and (3) a gauge-aware relative positional bias built from local tangent-plane angles rather than global axes. A logit-space \emph{SO(3)}-consistency regularizer, used only during training, further suppresses residual discretization effects. To benchmark robustness, we introduce Pose35, a variant of Stanford2D3D perturbed by random rotations within $\pm 35^\circ$, and evaluate under a full, arbitrary \emph{SO(3)} stress test. There, the baseline SphereUFormer collapses from 67.53 \emph{mIoU} on Pose35 to 25.26 under the full \emph{SO(3)} test, whereas SO3UFormer reaches 72.03 on Pose35 and retains 70.67 under the same test. Similarly, on a second real-world dataset (Matterport3D) for segmentation and on panoramic depth estimation, SO3UFormer remains essentially rotation-invariant while the gravity-anchored baseline again loses most of its accuracy. Code and models are available at https://github.com/zhuqinfeng1999/SO3UFormer.

cs.CV↗

ClassWise-CRF: Category-Specific Fusion for Enhanced Semantic Segmentation of Remote Sensing Imagery

We propose a result-level category-specific fusion architecture called ClassWise-CRF. This architecture employs a two-stage process: first, it selects expert networks that perform well in specific categories from a pool of candidate networks using a greedy algorithm; second, it integrates the segmentation predictions of these selected networks by adaptively weighting their contributions based on their segmentation performance in each category. Inspired by Conditional Random Field (CRF), the ClassWise-CRF architecture treats the segmentation predictions from multiple networks as confidence vector fields. It leverages segmentation metrics (such as Intersection over Union) from the validation set as priors and employs an exponential weighting strategy to fuse the category-specific confidence scores predicted by each network. This fusion method dynamically adjusts the weights of each network for different categories, achieving category-specific optimization. Building on this, the architecture further optimizes the fused results using unary and pairwise potentials in CRF to ensure spatial consistency and boundary accuracy. To validate the effectiveness of ClassWise-CRF, we conducted experiments on two remote sensing datasets, LoveDA and Vaihingen, using eight classic and advanced semantic segmentation networks. The results show that the ClassWise-CRF architecture significantly improves segmentation performance: on the LoveDA dataset, the mean Intersection over Union (mIoU) metric increased by 1.00% on the validation set and by 0.68% on the test set; on the Vaihingen dataset, the mIoU improved by 0.87% on the validation set and by 0.91% on the test set. These results fully demonstrate the effectiveness and generality of the ClassWise-CRF architecture in semantic segmentation of remote sensing images. The full code is available at https://github.com/zhuqinfeng1999/ClassWise-CRF.

cs.CV↗