arXiv subjects

Ning Liu

Publications and source records attributed to Ning Liu.

At least 19 recordsLinked to original sources

arXiv2026

CARE: Consistent Action-Response Execution for Human-Guided Robot Transport

Human-guided collaborative transport becomes interaction-critical when the operator can continue along a route segment that is not executable by the robot--payload system. In this regime, operator motion is not an exogenous reference: it simultaneously communicates a persistent route tendency and adapts to the robot's realized progress. A common modular formulation predicts one operator future from observation history, scores multiple robot proposals against that shared forecast, and then modifies the selected command for local feasibility. This can compare candidates under responses they would not induce and can condition prediction on an action that is never executed. We formulate local control over executable-action-indexed joint human--robot transitions and instantiate it in Consistent Action--Response Execution (CARE). Each nominal proposal or bounded alternative is first mapped to an executable action; that same action conditions the route-aligned operator-response prediction, the matched robot--operator rollout, and a fixed interaction utility for candidate selection. Bounded reciprocal refinement handles transient route--feasibility mismatch, whereas persistent progress loss replaces the nominal proposal with a progress-oriented recovery proposal while retaining the same downstream executable-action-conditioned evaluation. Across training and held-out maps, CARE preserves near-saturated performance when the indicated route remains locally usable and achieves the highest observed success and service availability when the operator can continue but the robot must detour. It also yields lower jerk-proxy values and measured controller latency than dynamic-window and Timed Elastic Band baselines. Factorial ablations distinguish reciprocal refinement from progress recovery, delayed-response tests probe temporal mismatch, and an indoor deployment demonstrates physical executability.

cs.RO↗

arXiv2026

VORL-EXPLORE: A Hybrid Learning Planning Approach to Multi-Robot Exploration in Dynamic Environments

Hierarchical multi-robot exploration decouples frontier allocation from local navigation, making systems brittle in dynamic environments. Without awareness of execution difficulty, allocators may send robots toward frontiers whose routes are congested or unsafe, causing bottlenecks, oscillatory replanning, and redundant coverage. We propose VORL-EXPLORE, a hybrid learning and planning framework that makes exploration execution-aware through execution fidelity, an estimate of local navigability and execution reliability. This fidelity score is used to penalize risky or crowded frontiers in a fidelity-coupled Voronoi allocation objective with inter-robot repulsion, and to arbitrate between global $A^*$ guidance and a reactive reinforcement learning policy during navigation. As a result, robots follow efficient global plans in reliable regions and rely more on safer local reactions in crowded or changing areas. The fidelity model is recalibrated using self-supervised pseudo-labels derived from recent progress and safety outcomes. Quantitative experiments on randomized dynamic grids and an ablation under severe traffic show higher success rates, shorter exploration lengths, and lower inter-robot coverage overlap. A Gazebo factory scenario is included as a deployment illustration. The source code is available at https://github.com/21ning/VORL-EXPLORE.

cs.RO↗

arXiv2026

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

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

cs.NE↗

arXiv2026

Direct Detection of Leptophobic Dark Matter with Electronic Collective Excitations

Some new-generation dark matter detection experiments are primarily designed to search for the dark matter-electron interactions, but they can also be utilized to probe models in which dark matter couples exclusively to nucleon via the quantum effects. The hadronic loop-induced interactions can directly excite plasmons in semiconductors, thereby providing an additional channel for detecting the leptophobic dark matter. In this work, we investigate plasmon excitations in silicon detectors induced by boosted dark matter and cosmic-ray up-scattering dark matter via the hadronic loop process. By analyzing the available experimental data, we derive new exclusion limits on the leptophobic dark matter-nucleon scattering cross section in the sub-MeV mass range.

hep-ph↗

arXiv2026

SERO: Learning from Executable Oracle Continuations for Long-Horizon Agentic Reinforcement Learning

Long-horizon agentic RL is constrained by the futures visited by the current policy. When rollouts fail to reach useful behaviors, improved credit assignment cannot provide supervision for states and actions that were never encountered. On-policy distillation (OPD) introduces teacher guidance on policy-induced states, and recent multi-turn variants improve how this guidance is scheduled, replayed, refined, or validated. However, they do not directly use the oracle-controlled future itself as the training trajectory. We introduce \textbf{State-Rooted Executable Rollout Optimization (SERO)}, which augments on-policy agentic RL with executable oracle continuations. SERO restores a state actually reached by the current policy and lets a training-only oracle continue interacting with the real environment, making the downstream observations and oracle actions themselves available for learning. It then applies \textbf{policy-reachability weighting}, which weights each oracle step by the frozen behavior policy's compatibility with the preceding oracle actions. This allows deeper supervision to become effective as the policy becomes more compatible with the oracle branch. The oracle loss is optimized separately from the base RL estimator and no oracle is used at inference time. We evaluate SERO on two challenging interactive-agent benchmarks, ALFWorld and WebShop, using Qwen2.5-1.5B-Instruct and Qwen2.5-7B-Instruct. On ALFWorld, learning the full executable continuation is substantially more effective than supervising only the root oracle action, while policy-reachability weighting further improves out-of-distribution performance.

cs.AI↗

arXiv2026

RoboAug: One Annotation to Hundreds of Scenes via Region-Contrastive Data Augmentation for Robotic Manipulation

Enhancing the generalization of robotic learning in diverse unseen environments remains a fundamental challenge. Existing approaches often rely on large-scale pretraining, which is labor-intensive and time-consuming, or semantic data augmentation methods that assume flawless upstream object detection in real-world scenarios. In this work, we propose RoboAug, a novel generative data augmentation framework that reduces reliance on large-scale pretraining and perfect visual recognition by requiring only a single image with bounding box annotations for dataset construction. Leveraging this minimal supervision, RoboAug employs pretrained generative models for precise semantic augmentation and introduces a plug-and-play region-contrastive loss to guide attention toward task-relevant regions, thereby enhancing generalization and task success rates. Extensive real-world experiments on UR-5e, AgileX, and Tian Gong 2.0 demonstrate that RoboAug consistently outperforms state-of-the-art augmentation baselines under background, distractor, and lighting shifts. Our project is available at https://x-roboaug.github.io/.

cs.RO↗

arXiv2026

Beyond Pairs: Your Language Model is Secretly Optimizing a Preference Graph

Direct Preference Optimization (DPO) aligns language models using pairwise preference comparisons, offering a simple and effective alternative to Reinforcement Learning (RL) from human feedback. However, in many practical settings, training data consists of multiple rollouts per prompt, inducing rich preference structure that pairwise DPO fails to exploit. Collapsing such data into independent pairs discards transitivity, introduces redundant or conflicting supervision, and can lead to unstable optimization. We propose Graph Direct Preference Optimization (GraphDPO), a structured generalization of DPO that operates over directed acyclic preference graphs induced by rollout rankings. GraphDPO encodes dominance relations as edges and optimizes a graph-structured Plackett--Luce-inspired objective that aggregates supervision over dominated neighborhoods, propagating the transitive structure of the input preferences while recovering standard DPO as a special case. GraphDPO naturally handles tied and incomparable responses through an equivalence-class construction: responses with identical preference signals form graph layers, incomparable responses share a layer without an edge, and absent edges contribute zero loss, preventing the spurious gradients that strict total-order objectives incur. Under this layered construction, GraphDPO maintains linear per-prompt complexity via efficient log-sum-exp aggregation. We further incorporate optional ground-truth anchoring by inserting verified solutions as dominant nodes and applying an annealed schedule that stabilizes early training while gradually relaxing oracle supervision. Experiments on reasoning and program synthesis tasks demonstrate superior performance, suggesting that graph-structured preference modeling is a scalable and robust alternative to pairwise and listwise alignment objectives.

cs.LG↗

arXiv2026

Quantum super Schur--Weyl duality and character formulas for mirabolic Hecke algebras

We establish a quantum super Schur--Weyl duality for the mirabolic Hecke algebra $H_n^{\mathrm{mir}}(q)$ with $q$ indeterminate and determine the corresponding bimodule decomposition. The centralizer of the mirabolic action is shown as a direct sum of homogeneous quantum Schur superalgebras. We describe the tensor-space annihilator and the resulting faithful quotient. When nonzero, the annihilator is generated by an explicit spectral idempotent attached to the smallest rectangle excluded by the hook condition. From the quantum super Schur--Weyl duality, we derive a super Frobenius formula for mirabolic Hecke algebras. This formula then yields a Murnaghan--Nakayama rule and Regev-type formulas for the irreducible characters. We also construct a super mirabolic RSK bijection from words in a super alphabet with an additional even letter to hook semistandard insertion tableaux together with recording pairs that distinguish the positions of the additional letter. This correspondence yields a Roichman formula for irreducible characters of $H_n^{\mathrm{mir}}(q)$. Finally, we establish a parameter-inversion isomorphism between $H_n^{\mathrm{mir}}(q^{-1})$ and the $q$-rook monoid algebra $R_n(q)$ and use it to transport the duality, annihilators, and character formulas in both directions. The transported tensor decomposition provides a representation-theoretic interpretation of hook and two-row character sums for $R_n(q)$.

math.RT↗

arXiv2026

SynAct: Synergistic Active Feedback Learning for LLM Agents

LLM agents executing multi-step workflows fail in two fundamentally different ways: recoverable slips, which further attempts can repair, and knowledge gaps, which retrying cannot resolve. Expert feedback can address the latter, but expert attention is scarce. The key challenge is therefore not whether to ask for feedback, but where feedback can change the outcome. Existing approaches either avoid expert feedback, allowing agents to reinforce their own errors, or allocate feedback at the task level without identifying consequential decisions. We present SynAct, an active learning framework that treats expert attention as the binding constraint and allocates feedback at the level of individual decisions. SynAct ranks decisions by the mutual information between the selected action and the terminal outcome across repeated rollouts, queries an expert at the highest-value decisions, and distills the resulting corrections into transferable rules that the agent can reuse without expert access at deployment. We theoretically characterize what this signal identifies under finite repeated sampling: with k rollouts and at least m observations per action, outcome-constant decisions have zero mutual information, while outcome-separating decisions have mutual information at least H2(m/k) bits, yielding exact separation for any threshold t in (0,H2(m/k)). This also explains why success-based acquisition can be blind when all rollouts fail: success is constant even when failed runs reach different outcomes. Across four agent benchmarks and four models spanning three families, SynAct improves held-out accuracy in all 16 evaluated settings, by +11.7pp on average. The same signal transfers to inference-time routing, where SynAct outperforms budget-matched cascade, uncertainty, and expert-judge routers in 15 of 16 settings, while a single targeted correction outperforms seven additional unaimed retries.

cs.LG↗

arXiv2026

Revisiting the Generalization of Neural Graph Edit Distance Models

Neural approaches to Graph Edit Distance (GED) have achieved strong results under standard within-dataset evaluation, but much less is known about how well these models transfer across graph collections. We conduct a systematic study of this problem using exact GED supervision across diverse graph datasets and a broad set of representative learning-based methods. Our results reveal a pronounced gap between within-collection performance and cross-collection transfer. Models that perform well on their training collections often lose this advantage when evaluated on structurally different data. Training on multiple source collections substantially improves zero-shot transfer and provides a better starting point when limited supervision is available for a new target collection. Further analysis shows that transfer behavior varies with the source--target direction and the structural characteristics of the collections involved. These findings suggest that conventional within-collection evaluation provides only a partial view of the generalization behavior of neural GED models and motivate broader evaluation across heterogeneous graph collections.

cs.LG↗

arXiv2026

Artificial intelligences and human scientists exhibit complementary strengths in theory building

We investigate the effectiveness of artificial intelligences (AI)-specifically large language models (LLMs)-relative to human scientists at high-level cognitive tasks in social science such as theory formulation, predictions of novel empirical results, and theory revision in response to new evidence. The research domain was academic discourse regarding gender and race inequality. Our findings, comparing 25 LLMs with 13 senior researchers and 60 doctoral scholars, reveal that the AIs outperformed most humans individually on most of the present tasks, while human theories were more diverse and exhibited greater gains in predictive accuracy from aggregation. AI-generated theories were more extensively elaborated, involving additional theoretical paths and latent variables, and were rated as higher quality than human theories by independent raters blinded to source. However, this theoretical complexity was in part ornamental, in that it was not associated with more accurate predictions about empirical patterns in data; in contrast, human scientists achieved greater predictive efficiency with simpler theories. The AIs were significantly more likely than human scientists to revise their theories to incorporate new evidence; human scientists updated their beliefs in a selective way that is sensitive to prior prediction errors. We speculate that the superior processing capacity of artificial intelligences makes them especially well-suited to tasks requiring grappling with complexity, but that the greater diversity of human ideas is essential to wise crowds and collective creativity.

cs.AI↗

arXiv2026

Energy and Angular-Momentum Redistribution in Hydrogen Migdal Ionization

MARVEL's first direct observation of Migdal ionization in neutron scattering marks an experimental milestone and opens a new avenue for probing electronic response to nuclear recoil. Hydrogen, both the simplest atom and a constituent of its molecular target, provides a controlled benchmark. Within the nonrelativistic sudden approximation, we compute its energy-differential and integrated ionization probabilities with the full recoil phase. At the recoil parameter $β\equiv v_N/(αc)=1.04$, the full-to-dipole spectral ratio rises from $0.43$ when the emitted-electron energy is $1\%$ of the hydrogen binding energy to $5.0$ when it is $2.3$ times that energy. Near $β=10$, partial waves with orbital angular momentum $\ell\ge3$ carry more than $90\%$ of the calculated continuum probability. An independent bound-state-closure evaluation verifies the absolute normalization: at eight recoil values, its ionization probabilities agree with the continuum-integrated results to relative discrepancies below $4.4\times10^{-7}$. These results extend the hydrogen dipole response to the fast-neutron regime and establish energy and angular-momentum redistribution as linked consequences of resolving the recoil phase across an atom.

physics.atom-ph↗

arXiv2026

On the stability of solutions to the 3-D anisotropic Navier-Stokes equations in the critical space

In this paper, we study the continuous dependence of solutions to the three-dimensional incompressible anisotropic Navier-Stokes equations $(ANS)$ with initial data in the critical space $\dot{B}^{0,\frac12}$. We prove that the data-to-solution map is continuous in this setting. Due to the critical regularity of the problem, the classical Bona-Smith argument \cite{BoSm} (see also the abstract result \cite{ABITZ}) does not appear to apply because a key ingredient (the Lipschitz continuity at lower regularity) does not appear to hold. As a consequence, we develop a new direct Fourier-weighted method within the Besov framework.

math.AP↗

arXiv2026

MVFA: A Multi-View Text-Guided Multimodal Fusion LLM Adapter for Sentiment Analysis and Emotion Recognition

Multimodal sentiment analysis and emotion recognition in conversations demand effective modeling of heterogeneous interactions across textual, acoustic, and visual modalities. Although large language models (LLMs) offer powerful language understanding, adapting them to multimodal affective computing remains challenging: full-model fine-tuning is computationally prohibitive, while many existing lightweight adapters fail to preserve rich textual cues during cross-modal fusion. To address these limitations, we propose the multi-view text-guided multimodal fusion adapter (MVFA), a parameter-efficient framework that augments frozen LLMs with strong multimodal reasoning capability. MVFA first constructs complementary text views via max pooling, mean pooling, and attention pooling; these views then guide cross-modal interactions with audio and visual features. The fused multimodal representations are subsequently compressed into a compact set of learnable pseudo-tokens through an Enhanced Q-Former Fusion Module. Using ChatGLM3-6B-base as the primary backbone, we further validate MVFA on LLaMA2-7B and Qwen3-8B to examine its portability across multiple frozen LLM backbones. MVFA is evaluated on three challenging datasets: CH-SIMS V2.0, MELD, and CHERMA. Experimental results demonstrate that MVFA achieves state-of-the-art performance on key metrics while updating only a small fraction of parameters. Specifically, it attains 84.62\% Acc2 and 84.59\% F1 on CH-SIMS V2.0, 67.36\% Acc and 66.03\% WF1 on MELD, and 74.66\% Acc on CHERMA. These findings establish multi-view text-guided fusion as an effective and scalable paradigm for parameter-efficient multimodal LLM adaptation in affective computing. The code is publicly available at https://github.com/Overwhelm1208/MVFA.

cs.AI↗

arXiv2026

SafeAtlas-VL: Beyond Binary Multimodal Safety with Large-Scale Data and Guard Models

Multimodal safety moderation requires distinguishing risks arising from visual content, user intent, and assistant behavior. Existing safeguards, however, are typically trained for a single judgment target and reduce safety assessment to a binary decision. Consequently, risk becomes difficult to compare across a multimodal interaction, and ambiguous cases are obscured. We introduce SafeAtlas-VL, a dataset of 1.5M training instances that places image-, request-, and response-level judgments on a five-level ordered scale. We curate a broad collection of safety-relevant data from both real-world and synthetic sources and apply a disagreement-aware annotation procedure. The resulting dataset spans 15 harm categories and 55 fine-grained subcategories, covering a broad range of multimodal safety scenarios. We also construct SafeAtlas-Bench, a held-out set of 5,000 instances for evaluating five-level predictions and continuous risk scores. Upon this dataset, we train the SafeAtlas Guard series of models via target-conditioned tuning for multimodal safety detection. Our models not only perform five-way classification of safety levels but also map safety to continuous scores through a soft cumulative ordinal head. Experimental results demonstrate that guard models trained on our dataset exhibit strong generalization: even without using the training sets of other benchmarks, they achieve competitive performance on the corresponding test sets. Notably, our 8B model attains the overall best performance, outperforming the previous SOTA by approximately 4% in F1 score. Code, data, and models are released to support further research. Warning: this paper contains example data that may be offensive, harmful, graphic, or disturbing.

cs.AI↗

arXiv2026

Tool-Making and Self-Evolving LLM Agents in Low-Latency Systems

Production LLM agents often waste latency and reliability by regenerating code for the same procedural steps on every request. We replace this inference-time coding loop with an agentic tool-making pipeline that compiles repeated SOP steps into validated, versioned tools before deployment. The tool-maker grounds synthesis in the live environment as it collects execution traces, observes backend schemas and values, generates candidate tools, and repairs them against labeled cases. At runtime, the production agent calls these tools directly and falls back to code generation only when needed. We deploy the approach in a Fulfillment Center alarm-triage system, where an agent diagnoses alarms against a 44-node SOP over heterogeneous metric backends. In production, tool calls reduce p50 latency by 42%. On 1,500 historical alarms, they reduce end-to-end error rate by up to 53% by suppressing run-to-run variance in repeated steps. Because tools return compact structured verdicts, they also enable a simpler direct-call architecture, reducing p50 latency by a further 62% in a controlled ablation. Versioned tools also improve auditability and expose specification gaps and upstream data drift. Our results show that agents that build and maintain their own tool libraries, a key element of self-evolving agents, can make industrial LLM systems faster, more reliable, and easier to operate.

cs.CL↗

arXiv2026

Eluna: An Agentic LLM System for Automating Warehouse Operations with Reasoning and Task Execution

Warehouse operations are governed by Standard Operating Procedures (SOPs) that encode complex, multi-system decision logic, which must be executed reliably under strict time constraints, yet LLM agents lack mechanisms to enforce procedural compliance and degrade under the context overload full SOP specifications introduce. We present Eluna, a production-deployed agentic system for reliable SOP execution. Eluna is a graph-guided, multi-agent framework that encodes SOPs as directed acyclic graphs with progressive disclosure and delegates independent tasks to parallel sub-agents, each with persistent code execution and live data access. To meet production latency and accuracy needs, we use asymmetric episodic distillation where a strong teacher is improved through episodic error memories, then a smaller student is fine-tuned on the corrected trajectories with memory stripped, internalizing corrections without inference-time overhead. On a 13-task benchmark and two production applications, our fine-tuned models match or exceed their teacher, beat all larger off-the-shelf baselines, and reach 94% expert agreement on the ticket processing application.

cs.LG↗

arXiv2026

VTONQA: A Multi-Dimensional Quality Assessment Dataset for Virtual Try-on

With the rapid development of e-commerce and digital fashion, image-based virtual try-on (VTON) has attracted increasing attention. However, existing VTON models often suffer from artifacts such as garment distortion and body inconsistency, highlighting the need for reliable quality evaluation of VTON-generated images. To this end, we construct \textbf{VTONQA}, the first multi-dimensional quality assessment dataset specifically designed for VTON, which contains 8,132 images generated by 11 representative VTON models, along with 24,396 mean opinion scores (MOSs) across three evaluation dimensions (\textit{i.e.}, clothing fit, body compatibility, and overall quality). Based on VTONQA, we benchmark both VTON models and a diverse set of image quality assessment (IQA) metrics, revealing the limitations of existing methods and highlighting the value of the proposed dataset. We believe that the VTONQA dataset and corresponding benchmarks will provide a solid foundation for perceptually aligned evaluation, benefiting both the development of quality assessment methods and the advancement of VTON models. The dataset we proposed in this paper is publicly available at: https://huggingface.co/datasets/weixiny0408/vtonqa.

cs.CV↗