arXiv subjects

Jiahao Liu

Publications and source records attributed to Jiahao Liu.

At least 19 recordsLinked to original sources

arXiv2026

Disentangling Spin and Orbital Contributions in the Search for Altermagnetic Terahertz Conversion in RuO2

Rutile RuO2 has emerged as a prominent candidate altermagnet, yet the microscopic origin of its reported charge-angular-momentum conversion remains unsettled. Recent theoretical and experimental advances have highlighted a potentially dominant role of orbital transport, raising a fundamental ambiguity between spin- and orbital-mediated conversion in RuO2. Here, we report the first experimental disentanglement of the spin- and orbital-to-charge conversion channels in RuO2 heterostructures using terahertz emission spectroscopy. We identify Fe as a predominantly spin-current injector, whereas Ni and permalloy predominantly inject orbital currents, revealing a substantial orbital contribution to ultrafast terahertz emission. Under controlled spin-current injection, we find no evidence for inverse altermagnetic spin-splitting effect. Instead, the observed anisotropy is consistent with an anisotropic inverse spin Hall response that is strongly conditioned by the near-interface environment. Under orbital-current injection, the anisotropy is markedly weaker, indicating that orbital-to-charge conversion is largely isotropic in RuO2. Our results establish orbital transport and interface-conditioned spin Hall conversion as key to terahertz emission from RuO2, underscoring the need to disentangle spin, orbital, and interfacial contributions before assigning anisotropy to altermagnetism.

cond-mat.mes-hall↗

arXiv2026

StreamMemBench: Streaming Evaluation of Agent Memory for Future-Oriented Assistance

A central role of personal-agent memory is to turn stored information and prior interactions into future-oriented assistance. In daily use, useful cues come from what the agent observes and how the user interacts with the agent, and the agent must carry them forward from the current request to similar future tasks. Existing memory benchmarks usually test dialogue recall or task improvement in isolation, leaving the trajectory from streaming observations to later assistance largely untested. We introduce StreamMemBench, a streaming benchmark that constructs a two-step task sequence around each evidence anchor from EgoLife egocentric streams. The initial task tests evidence use, while the follow-up task tests whether feedback and interaction experience are reused. Four metrics diagnose evidence recall, initial evidence use, feedback incorporation, and follow-up reuse. Experiments with eight memory systems across two backbones show that current systems often fail to use observed evidence or turn feedback into reliable follow-up behavior, even when evidence is stored or feedback is incorporated locally. StreamMemBench is publicly available at https://github.com/landian60/StreamMemBench.

cs.AI↗

arXiv2026

Quantifying Ultrafast Orbital Pumping and Nanometre-Scale Orbital Diffusion in Magnetic Heterostructures

Quantifying ultrafast orbital pumping and transport remains challenging because spin and orbital currents contribute coherently to terahertz emission. Here, we combine spintronic terahertz emission spectroscopy with wedge-sample thickness control to disentangle these contributions in Fe, Co and Ni heterostructures with Ta, W and Nb. Their opposite spin Hall and orbital Hall polarities enable quantitative determination of relative pumping strengths and propagation lengths. We find that orbital pumping increases from Fe to Ni, with inferred orbital-to-spin current ratios reaching several tens of percent in Ni, exceeding current microscopic predictions. Remarkably, Nb exhibits an orbital diffusion length of ~1.5 nm, substantially longer than in Ta and W and beyond the sub-nanometre scale predicted by recent orbital-relaxation calculations. These findings provide quantitative experimental benchmarks for ultrafast orbital pumping and transport.

cond-mat.mes-hall↗

arXiv2026

A periodic $\bar q$ equation for form factors

MHV form factors of the chiral stress-tensor multiplet in planar $\mathcal N=4$ super Yang-Mills theory are dual to periodic null Wilson loops. We propose a periodic $\bar q$ equation for the MHV and higher-helicity sectors at vanishing fermionic monodromy. The equation relates kinematic derivatives of $\mathrm N^k\mathrm{MHV}$ form factors to collinear integrals of $\mathrm N^{k+1}\mathrm{MHV}$ form factors with one additional external particle, providing a recursive relation in loop order. In periodic momentum supertwistors, it has the same collinear-integral form as the two dotted components of the amplitude $\bar Q$ equation, with coefficient $Γ_{\rm cusp}/4$, despite the absence of ordinary dual conformal invariance at fixed operator momentum. For parity-even scalar MHV remainders, combining the equation with its parity image supplies all directions of the kinematic differential. We verify the equation analytically for the one-loop four-point NMHV form factor and reconstruct the two-loop three-point MHV remainder at function level. Further two-loop MHV tests give numerical agreement with derivatives of the known four-point remainder at a fixed kinematic configuration, and for five points, all six chiral derivatives agree exactly at symbol level for a one-parameter family of kinematics. The structure of the collinear source further gives all-loop predictions for MHV last entries, consistent with known results and form a space of dimension $n(2n-5)$. The periodic Wilson-loop picture also motivates an all-loop derivation. Using the form-factor OPE, we extract the one-loop correction to the fermion energy from the logarithmic source and recover its vanishing at zero momentum.

hep-th↗

arXiv2026

TRACT: Temporally Routed Action Chunks with Chronological Phase Authority for Contact-Rich Manipulation

Action chunking shortens the effective decision horizon of robot imitation learning by predicting multiple future actions, while conventional phase conditioning describes the current control instant. When a predicted horizon crosses a procedural boundary, assigning the current phase to the entire chunk creates a structural temporal mismatch. We present TRACT, which factorizes phase-structured action chunking into an accepted current phase and a single CURRENT-to-NEXT boundary inside the future horizon. A task-local graph constrains chronological phase authority, and a cumulative boundary distribution monotonically routes future queries through phase-specific query and action paths. For contact execution, a causal response-deficit integrator compares policy intent with ACK-eligible subsequent motion, accumulates arm compensation when directional response is suppressed, and decays after confirmed recovery. Across six real-robot variants with ten trials each, full TRACT achieves 10/10 full-sequence success, 99.00 [88.75, 100.00]% median [min, max] wipe completion, zero observed phase ambiguity, and zero stalls. Under the current complete method package and evaluation setting, the routed representation obtains better observed task results than the flat package (6/10 vs. 3/10 success; 77.08% vs. 8.03% median wipe completion). Chronological authority reduces observed phase ambiguity from 8/10 to 0/10, and response integration reduces stalls from 4/10 to 0/10. The package comparison does not isolate routing from other generator-package differences.

cs.RO↗

arXiv2026

From Approval to Abuse: Privilege Escalation Attacks on Harness-Protected Coding Agents

Coding agents use execution harnesses to constrain the consequences of prompt injection by enforcing permission boundaries. These systems deterministically intercept any out-of-sandbox action and prompt the user for approval. To reduce approval fatigue, users can approve allow-rules proposed by the agent, after which any future action matching the rule is allowed without a prompt. In this work, we show that this mechanism introduces a new risk: untrusted context can influence the permissions presented for approval, with consequences extending beyond the immediate task. We investigate privilege escalation targeting this permission-update process. Specifically, adversarial context planted in a repository can reliably steer an agent performing a routine task into requesting a plausibly justified, yet dangerously broad allow-rule. If the user approves the request, the saved rule can effectively blind the harness, enabling adversaries to execute malicious actions that would otherwise require approval, without triggering human oversight. Across production platforms, including Claude Code and Codex CLI, we demonstrate that agents successfully elicit the target allow-rules in 58% of experimental instances. Once acquired, these rules exhibit a massive semantic blast radius, covering 56.9% of category-specific capability test cases. To further quantify the downstream risk, we show that conditional on rule approval, our exploit achieves verified, silent goal completion in 47 of the 58 eligible instances (81.0%), undermining the safety promise of human-in-the-loop oversight.

cs.AI↗

arXiv2026

WoVR+: World Model as Reliable Simulator for Versatile VLA Policy Optimization with RL

Action-conditioned world models can reduce costly real-robot interaction in reinforcement learning for Vision-Language-Action models (VLA-RL), but hallucinations can mislead policy optimization. Extending them to richer VLA interfaces and long-horizon tasks raises challenges in spatiotemporal and cross-modal consistency, exploration-reliability balance in RL, and compounding rollout errors. We present WoVR+, extending WoVR's hallucination-aware WM-RL framework by redesigning the learned environment and imagined interaction protocol: A VLA-interface-aligned world model provides multi-view prediction and proprioceptive updates while mitigating hallucinations. Chunk-wise noise preserves flow-based policy exploration while limiting prediction-degrading action jitter. For long-horizon tasks, stage-local imagined training combines stage-entry resets, stage-specific rewards, local horizons, and controllable sampling to shorten generated prefixes and balance training across stages. WoVR+ improves success over SFT by 21.09% and 16.7% on RoboTwin and 6 real-robot tasks. Long-horizon phone-packaging success rises from 30% to 75%. Videos are available at https://youmo445.github.io/WoVR_Plus.github.io/.

cs.RO↗

arXiv2026

To Cache or Not to Cache: Understanding and Mitigating Safety Incompatibility in Cross-Model KV Cache Reuse

Cross-model KV cache sharing reduces prefill computation by allowing a receiver model to reuse prefix KV states generated by a related sender model. Existing methods optimize such reuse for latency and generation quality, but its safety implications remain unclear. In this work, we systematically evaluate this issue across diverse sender-receiver pairs and reuse configurations. We refer to deviations from the receiver's native safety behavior as safety incompatibility. Our study shows that both the direction and magnitude of safety incompatibility depend on the model pair and reuse configuration, with value reuse causing substantially larger deviations than key reuse across all evaluated pairs. Furthermore, we propose VReparo, a technique that learns per-layer and per-KV-head affine maps offline to align reused sender value tensors with their receiver-native counterparts before decoding. Across two model pairs under full KV reuse and the state-of-the-art selective reuse setting, VReparo reduces the average HarmBench attack success rate (ASR) from 57.28% to 39.21% and the average StrongREJECT score from 0.4215 to 0.2163 upon our benchmarks, while retaining an average Time to First Token (TTFT) reduction of 85.62%.

cs.LG↗

arXiv2026

Broadband LEO Satellite Constellations for Next-Generation Navigation: Potentials, Enabling Technologies, and Challenges

The rapid deployment of broadband low Earth orbit (LEO) constellations provides new opportunities for next-generation navigation. Compared with conventional global navigation satellite systems (GNSS), broadband LEO systems provide stronger received signals, wider bandwidths, rapidly varying satellite geometry, and larger constellation scales, offering new capabilities for positioning, navigation, and timing (PNT). Meanwhile, their communication-oriented waveforms, high dynamics, and resource-constrained architectures introduce new challenges that are fundamentally different from those in GNSS. This article provides a comprehensive overview of broadband LEO constellations for next-generation navigation, covering their fundamental characteristics, representative use cases, key enabling technologies, and open challenges. We first compare broadband LEO systems with conventional GNSS and summarize the main technical routes for LEO-based PNT. We then discussed the emerging use cases in GNSS-challenged, multipath-dominant, and integrated communication and navigation (ICAN) scenarios. Key enabling technologies including waveform design, high-dynamic signal processing, and constellation resource scheduling are further reviewed. As a representative case study, the navigation-oriented optimization of the Starlink primary synchronization sequence demonstrates that subcarrier power allocation can improve multipath resistance while maintaining ranging accuracy. Finally, several challenges and future directions of broadband LEO-based navigation are discussed, aiming to provide reference for future researches.

eess.SP↗

arXiv2026

InCoM: Intent-Driven Perception and Structured Coordination for Mobile Manipulation

Mobile manipulation is a fundamental capability for general-purpose robotic agents, requiring both coordinated control of the mobile base and manipulator and robust perception under dynamically changing viewpoints. However, existing approaches face two key challenges: strong coupling between base and arm actions complicates control optimization, and perceptual attention is often poorly allocated as viewpoints shift during mobile manipulation. We propose InCoM, an intent-driven perception and structured coordination framework for mobile manipulation. InCoM infers latent motion intent to dynamically reweight multi-scale perceptual features, enabling stage-adaptive allocation of perceptual attention. To support robust cross-modal perception, InCoM further incorporates a geometric-semantic affinity refinement mechanism that enhances multimodal correspondence. On the control side, we design a decoupled coordinated flow matching action decoder that explicitly models bidirectional base-arm interactions, alleviating optimization difficulties caused by control coupling. Under matched non-privileged observations and task-specific training data, InCoM improves mean success over DSPv2 by 33.0, 26.1, and 23.6 percentage points on SetTable, TidyHouse, and PrepareGroceries, respectively. Furthermore, its effectiveness is consistently validated in real-world mobile manipulation tasks, where InCoM maintains a superior success rate over existing baselines.

cs.RO↗

arXiv2026

WALA Learning Executable Latent Actions from Action-Labeled Demonstrations and Action-Free Videos

Human videos provide rich information about object manipulation at a scale difficult to reproduce with robots, but often lack action annotations that can directly supervise robot policies. Realizing this potential requires distinguishing interaction-relevant changes from static appearance and connecting the learned interaction knowledge to executable robot actions. We present WALA, a framework named for World- and Action-supervised Latent Actions. Its semantic-geometric latent action model (LAM) learns latent actions by encoding and predicting semantic and geometric changes, emphasizing interactions over static appearance while retaining spatial detail. During robot policy training, the LAM encoder and decoder provide guidance through distillation and visual prediction, respectively. Together with robot action supervision, these signals shape policy-generated latent actions to support world prediction and capture the information needed to directly generate executable robot actions. Because LAM supervision requires no action labels, WALA enables co-training on action-labeled robot demonstrations and task-relevant action-free videos. In simulation experiments, our method achieves 92.33% average success on RoboTwin and 75.2% on RoboCasa-GR1-Tabletop, exceeding the previous state of the art on RoboCasa by 5.0 percentage points. WALA also shows strong real-robot performance, further enhanced by task-relevant action-free human videos that improve data efficiency and support target-task zero-shot and few-shot transfer.

cs.RO↗

arXiv2026

World In Your Hands: A Large-Scale and Open-Source Ecosystem for Learning Human-Centric Manipulation in the Wild

We introduce World In Your Hands (WIYH), a large-scale open-source ecosystem comprising over 1,000 hours of human manipulation data collected in-the-wild with millimeter-scale motion accuracy. Specifically, WIYH includes (1) the Oracle Suite, a wearable data collection kit with an auto-labeling pipeline for accurate motion capture; (2) the WIYH Dataset, featuring over 1,000 hours of multimodal manipulation data across hundreds of skills in diverse real-world scenarios; and (3) extensive annotations and benchmarks supporting tasks from perception to action. Furthermore, experiments based on the WIYH ecosystem show that integrating WIYH's human-centric data improves robotic manipulation success rates from 8% to 60% in cluttered scenes. World In Your Hands provides a foundation for advancing human-centric data collection and cross-embodiment policy learning. All data and hardware design will be open-source.

cs.RO↗

arXiv2026

CIBuzzBench: A Benchmark for Cross-Lingual Understanding of Chinese Internet Buzzwords

Chinese social media has generated a vast and continually evolving lexicon of internet buzzwords whose meanings are often non-literal and deeply rooted in local cultural and pragmatic contexts. Existing research has primarily focused on interpreting these buzzwords within Chinese, leaving largely unexplored whether LLMs can transfer such culturally grounded knowledge across languages and accurately convey the intended meanings in English. This cross-lingual capability is also critical for safety, as harmful expressions may obscure their offensive content through culture-specific homophony, euphemism, irony, or coded language. In this paper, we investigate the ability of advanced LLMs to understand Chinese internet buzzwords across languages. To this end, we introduce CIBuzzBench, the first benchmark for cross-lingual Chinese-to-English understanding of Chinese internet buzzwords. CIBuzzBench comprises 3,001 Chinese internet buzzwords annotated with English meaning explanations, English equivalents, category labels, and harmfulness labels. Based on these annotations, we design three evaluation tasks: Meaning Explanation, Cross-lingual Equivalent Matching, and Culturally Grounded Harmfulness Detection. We evaluate representative state-of-the-art proprietary and Chinese LLMs under both English- and Chinese-prompting settings. Our results show that LLMs continue to struggle with the cross-lingual understanding of Chinese internet buzzwords, particularly in fine-grained non-literal interpretation, robust equivalent matching under option perturbations, and calibrated harmfulness detection. These findings highlight the persistent challenges posed by culturally grounded language phenomena for multilingual LLMs and safety-oriented evaluation. The dataset and code are available at https://github.com/SuperYFan/CIBuzzBench.

cs.CL↗

arXiv2026

Continuous Manifold-Decomposed Impedance Retargeting for Contact-Rich Imitation Learning

CMDIR extends Manifold-Decomposed Impedance Retargeting (MDIR) to transform fixed-impedance demonstrations into continuous variable-impedance controllers, which can also serve as structured supervision for imitation learning. Continuous Task-Manifold Impedance Representation (TMIR) pairs an evolving task frame with controller instructions. Demo-relative Compromise dynamics retain moving-basis transport and control/physical metric mismatch, yielding displacement, reaction-impulse, and perturbation-sensitivity criteria. Quality-to-Fast automatically compiles a solver structure from development paths within a predefined finite space, re-instantiates that structure for each demonstration, and certifies the resulting candidate by multi-resolution evaluation. Across 225 retargeted-controller trials in three real contact tasks, full CMDIR improves mean task-proxy retention and reduces mean pose deviation, force fluctuation, and peak force relative to discrete MDIR. FastMPO achieves a $5.8$--$9.4\times$ speedup over C-MPO with comparable closed-loop outcomes. Downstream experiments demonstrate learnability of the complete TMIR supervision interface; lower force fluctuation and peak force are observed among successful executions, while completion reliability remains uneven across tasks and environments.

cs.RO↗

arXiv2026

Dynamic Feature-Embedding Communication via Codebook Distillation for Federated Recommendation

Federated recommendation systems commonly protect user privacy by keeping user parameters on local devices, while exchanging item parameters for collaborative model training. However, such item parameters usually model items independently and suffer from both efficiency and effectiveness challenges, making communication costs grow with the item space and limiting cross-item generalization and robustness to noisy feedback. To address these limitations, we propose to model items via shared latent feature embeddings for communication. Residual Quantization (RQ) provides a natural way to instantiate this communication by representing each item with a short sequence of discrete code IDs, i.e., Semantic IDs (SIDs). However, directly applying centralized and static RQ-based recommendation to federated learning is non-trivial due to 1) private and biased historical interactions and 2) evolving collaborative information. We propose RQFedRec, an RQ-based federated recommendation framework for dynamic feature-embedding communication. To construct globally aligned codebooks without accessing private interactions, RQFedRec introduces an information distillation module. Each client first learns item embeddings that encode local collaborative information from private interactions, and then distills such information into feature-indexed codebooks under globally shared SIDs, making sparse and biased local signals more compatible with server aggregation. To adapt to evolving collaborative information, RQFedRec introduces a self-refining SID update module that dynamically refines global SID assignments from aggregated codebooks. Extensive experiments demonstrate that RQFedRec improves recommendation performance and reduces communication costs without relying on semantic information, while further benefiting from public semantics when available. Code is available at https://github.com/Mingzhe-Han/RQFedRec.

cs.IR↗

arXiv2026

FedHUR: Learning Hierarchical Utility-Guided Client Relations for Personalized Federated Recommendation

Federated recommendation enables collaborative model training while keeping user interaction data on local clients. A central problem in federated recommendation is how to aggregate useful information across clients for personalized recommendation. Existing personalized aggregation methods usually construct client relations from predefined parameter-based assumptions, such as parameter similarity or complementarity, and use these relations to determine aggregation weights. However, such methods construct a single global relation, which is insufficient to capture the hierarchical and multi-granularity nature of user relations in recommendation. Moreover, these predefined relations cannot directly reflect whether the related clients can improve prediction performance after aggregation. To address these limitations, we propose FedHUR, a federated recommendation framework for learning hierarchical utility-guided client relations. FedHUR takes item-item filters as the object for relation construction and aggregation. Specifically, it first aggregates and clusters each client's local information to obtain global hierarchical information. Each client computes hierarchical utility signals based on its local information and the global hierarchical information, indicating which collaborative information is useful for improving its prediction. The server uses these utility signals to retrieve clients that are useful to that client for further personalized aggregation. Extensive experiments on five real-world datasets show that FedHUR consistently outperforms existing federated recommendation baselines, demonstrating the effectiveness of hierarchical utility-guided client relation learning. Code is available at https://github.com/Mingzhe-Han/FedHUR.

cs.IR↗

arXiv2026

Form Factor Alphabets and Antipodal Duality from Cluster Algebras

We develop a cluster-algebraic description of symbol alphabets for chiral stress-tensor form factors in planar $\mathcal N=4$ super-Yang--Mills theory. Motivated by unfolding the three-point $C_2$ quiver, we propose a periodic framework for $n$-point form-factor alphabets based on folding a $\mathrm{Gr}(4,2n)$ quiver with two periods of frozen nodes. At four points, the folding leads to the $\widetilde{\mathrm{VI}}$ cluster algebra, which has infinitely many cluster variables. Tropical truncation selects a finite set of rational coordinates and four physical limit rays. The $A_1^{(1)}$ mutation sequences approaching these rays generate the four square roots and the corresponding algebraic-letter spaces. Within the resulting finite set of rational candidates, antipodal closure excludes precisely eight additional letters. On the parity-preserving surface, the antipodal map is realized, after a further folding, by two commuting mutations followed by a relabelling. Furthermore, we find an alternative quiver within the same mutation class that becomes self-antipodal after parity folding, mirroring the antipodal self-duality of the four-point MHV form factor. Its double- and triple-collinear boundaries reproduce, respectively, the $C_2$ algebra of the three-point form factor and the $A_3$ algebra of the six-point amplitude; after folding the latter to $C_2$, the antipodal map exchanges the two $C_2$ boundaries.

hep-th↗

arXiv2026

ElderBench: Benchmarking Autonomous Mobile Agents for Older Adults

While autonomous mobile agents hold great potential for assisting older adults with smartphone usage, existing GUI benchmarks mainly rely on explicit, goal-oriented instructions and rarely capture the naturally occurring language patterns of older users, such as indirect speech, referential ambiguity, and under-specified requests. This mismatch between benchmark instructions and real-world elderly interactions may hinder reliable agent deployment. To address this gap, we present ElderBench, the first benchmark for evaluating mobile GUI agents in authentic elderly-oriented scenarios. ElderBench is constructed from 249 naturally elicited smartphone tasks collected from older adults across 20 applications. We first characterize the linguistic divergence between elderly instructions and existing GUI benchmark instructions from syntactic, semantic, and pragmatic perspectives. We then evaluate mainstream GUI agents and Vision-Language Models under both online and offline settings, revealing substantial performance degradation when handling elderly-oriented instructions. Through controlled instruction normalization, failure analysis, and fine-grained linguistic feature analysis, we further identify how elderly-specific language patterns contribute to agent failures. Our findings provide actionable design insights toward more adaptive, interpretable, and age-inclusive GUI agents for older adults.

cs.AI↗