Searcharxiv⌕ Search

arXiv subjects

Gang Chen

Publications and source records attributed to Gang Chen.

At least 37 records · Page 2Linked to original sources
arXiv2026

From Context to Skills: Can Language Models Learn from Context Skillfully?

Many real-world tasks require language models (LMs) to reason over complex contexts that exceed their parametric knowledge. This calls for context learning, where LMs directly learn relevant knowledge from the given context. An intuitive solution is inference-time skill augmentation: extracting the rules and procedures from context into natural-language skills. However, constructing such skills for context learning scenarios faces two challenges: the prohibitive cost of manual skill annotation for long, technically dense contexts, and the lack of external feedback for automated skill construction. In this paper, we propose Ctx2Skill, a self-evolving framework that autonomously discovers, refines, and selects context-specific skills without human supervision or external feedback. At its core, a multi-agent self-play loop has a Challenger that generates probing tasks and rubrics, a Reasoner that attempts to solve them guided by an evolving skill set, and a neutral Judge that provides binary feedback. Crucially, both the Challenger and the Reasoner evolve through accumulated skills: dedicated Proposer and Generator agents analyze failure cases and synthesize them into targeted skill updates for both sides, enabling automated skill discovery and refinement. To prevent adversarial collapse caused by increasingly extreme task generation and over-specialized skill accumulation, we further introduce a Cross-time Replay mechanism that identifies the skill set achieving the best balance across representative cases for the Reasoner side, ensuring robust and generalizable skill evolution. The resulting skills can be plugged into any language model to obtain better context learning capability. Evaluated on four context learning tasks from CL-bench, Ctx2Skill consistently improves solving rates across backbone models.

cs.AI↗
arXiv2026

Connection-topology--dependent energy transport and ergotropy in quantum battery networks with reciprocal and nonreciprocal couplings

The realization of scalable quantum battery architectures requires concern not only with how much energy can be stored, but also with how energy is transported, distributed, and converted into extractable work across connected battery nodes. While previous studies mainly focused on collective charging in multi-cell quantum batteries, the topology-dependent transport law and the corresponding work-oriented performance of quantum battery networks remain largely unexplored. In this work, we investigate quantum battery networks with engineered reciprocal and nonreciprocal couplings and compare different connection topologies, including cascaded and parallel architectures, within a unified transport framework. In the nonreciprocal regime, the optimal coupling follows distinct scaling laws for the two connection topologies, namely $J_{\rm op}^{c}\propto N$ for cascaded transport and $J_{\rm op}^{p}\propto N^{-1/2}$ for parallel charging in the large-$N$ limit. In reciprocal cascaded networks, a parity-dependent spectral response produces an odd-even transport effect that is absent in the nonreciprocal and parallel configurations. We further analyze the role of thermal and squeezed reservoirs and show that thermal noise mainly increases passive energy, whereas squeezing enhances ergotropy and thus the useful fraction of stored energy. These results shift the emphasis from charging enhancement to transport engineering and provide architecture-level design principles for quantum battery networks.

quant-ph↗
arXiv2026

Risk-Aware Belief Control Barrier Functions over Random Finite Sets

Ensuring robot safety in unknown, dynamic environments is a fundamental requirement. It involves inferring the states of an unknown and time-varying number of moving objects from noisy, incomplete measurements. We address safe control under the induced multi-object state uncertainty with a risk-aware belief control barrier function (BCBF) framework. The uncertainty is captured by a random finite set (RFS) belief, estimated by a sequential Monte Carlo probability hypothesis density (SMC-PHD) filter that represents it with a set of particles. Building directly on these particles, we construct a nonsmooth BCBF, establish forward invariance of the safe set under continuous prediction, and derive an explicit condition under which discrete updates preserve safety. Simulation and real-world underwater experiments demonstrate the effectiveness and efficiency of the proposed approach.

cs.RO↗
arXiv2026

Data Preservation in High Energy Physics: Global Report 2026

This document summarizes the contributions to the 5th DPHEP workshop March 5-6, 2026, CERN, and reflects the advancements since 2024, as well as future milestones and tendencies. Impressive progress in HEP data preservation is observed. Legacy data revival was showcased through successful reanalysis of archived data using contemporary methods, demonstrating the long-term scientific value of preservation. Sustainability challenges were noted, emphasizing the need for long-term funding and institutional support to maintain data preservation infrastructure, particularly for legacy experiments transitioning to archival modes. Innovative transverse projects display constant progress towards common technologies for a robust and transferrable DP. In particular, there is a clear shift toward automation, with increasing use of AI and machine learning for data curation, metadata extraction, and workflow optimization. Open science momentum is growing, with wider adoption of FAIR principles and open data policies, and experiments committing to public releases.

hep-ex↗
arXiv2026

Purified OPSD: On-Policy Self-Distillation Without Losing How to Think

On-policy self-distillation (OPSD) has emerged as a promising paradigm for improving LLM reasoning, where a privileged teacher with access to reference solutions provides token-level supervision on the student's own generated trajectories. However, we find that OPSD consistently fails on long chain-of-thought (long-CoT) reasoning models, yielding at best marginal gains while destabilizing the reflective reasoning capability these models depend on. Through a novel decomposition of the teacher's supervision signal, we identify the root cause: the teacher's supervision is dominated by a reference-induced component that drives rote memorization of reference-specific shortcuts, while the question-conditioned, inference-transferable component is ignored or actively opposed. Based on this diagnosis, we propose a two-step solution. First, we construct a reference-only teacher (the same model conditioned on the reference without the question) to isolate the non-transferable component of the supervision signal; the residual after subtracting this component captures the question-conditioned, inference-transferable correction. Second, we use pointwise mutual information (PMI) as the mechanism to transform this residual into a well-formed PMI target distribution that the student can directly distill from, filtering out the reference-induced shortcut. Experiments on four long-CoT models across two datasets demonstrate consistent improvements over both the base model and standard OPSD, while preserving the models' natural epistemic behavior throughout training.

cs.AI↗
arXiv2026

Enhancing Nonreciprocity through Squeezing-Induced Symmetry Breaking

Reservoir engineering enables unidirectional energy and signal flow. We establish squeezing-induced symmetry breaking between two cavities as a guiding principle for exponentially amplifying reservoir-mediated nonreciprocity. Rather than a simple scaling of the coupling, this mechanism strategically redistributes the squeezing resources to relax experimental requirements, as single-cavity squeezing alone demands a much larger squeezing strength. Moreover, reservoir squeezing does not alter the system symmetry, but reshapes the noise correlations and thereby changes the system dynamics. The proposed mechanism improves the performance of the quantum battery by several orders of magnitude, including stored energy, charging power, and ergotropy, with the analytical expressions provided. Extending to the optical isolation, we observe a second-order exponential enhancement of the output signal. Our results open a new avenue for nonreciprocal quantum information processing and nonreciprocal quantum device design.

quant-ph↗
arXiv2026

Verification-Gated Agentic Mission-State Governance for Intelligent Industrial Multi-Robot Systems

Agentic artificial intelligence is increasingly used to decompose industrial tasks, propose robot actions, and adapt execution plans in dynamic cyber-physical environments. However, autonomous proposal generation alone does not guarantee that multi-robot industrial systems preserve task dependencies, resource ownership, safety holds, or repair boundaries during long-horizon execution. This paper introduces a verification-gated agentic mission-state governance framework for intelligent industrial multi-robot systems. The framework maintains two synchronized state objects: an evolving task forest for persistent hierarchy, delayed grounding, and repairable substructures; and a governed blackboard for online execution state, robot traces, resource locks, world beliefs, proposals, verification records, and scene-temporary constraints. From each forest--blackboard snapshot, a derived execution coupling topology exposes cross-branch dependencies for proposal verification, parallel-commit eligibility, and bounded repair. Candidate assignments, repairs, deferrals, and constraint updates may be generated by heuristic, optimization, or agentic reasoning modules, but they can update the committed mission state only after deterministic verification and atomic commit. We evaluate the framework in an indoor factory multi-robot scenario, 30-seed remote-construction stress benchmarks, structural ablations, and scalability probes. The results show improved verified and safety-audited mission-state progress with fewer invalid commitments, lock conflicts, duplicate assignments, abandoned nodes, and disruptive repairs under modeled mission predicates. The study positions agentic AI as a proposal-generating layer governed by inspectable mission-state verification rather than as an unchecked execution authority.

cs.RO↗
arXiv2026

Capturing Context-Aware Route Choice Semantics for Trajectory Representation Learning

Trajectory representation learning (TRL) aims to encode raw trajectory data into low-dimensional embeddings for downstream tasks such as travel time estimation, mobility prediction, and trajectory similarity analysis. From a behavioral perspective, a trajectory reflects a sequence of route choices within an urban environment. However, most existing TRL methods ignore this underlying decision-making process and instead treat trajectories as static, passive spatiotemporal sequences, thereby limiting the semantic richness of the learned representations. To bridge this gap, we propose CORE, a TRL framework that integrates context-aware route choice semantics into trajectory embeddings. CORE first incorporates a multi-granular Environment Perception Module, which leverages large language models (LLMs) to distill environmental semantics from point of interest (POI) distributions, thereby constructing a context-enriched road network. Building upon this backbone, CORE employs a Route Choice Encoder with a mixture-of-experts (MoE) architecture, which captures route choice patterns by jointly leveraging the context-enriched road network and navigational factors. Finally, a Transformer encoder aggregates the route-choice-aware representations into a global trajectory embedding. Extensive experiments on 4 real-world datasets across 6 downstream tasks demonstrate that CORE consistently outperforms 15 state-of-the-art TRL methods, achieving an average improvement of 9.20\% over the best-performing baseline. Our code is available at https://github.com/caoji2001/CORE.

cs.CV↗
arXiv2026

HARD-KV: Head-Adaptive Regularization for Decoding-time KV Compression

Long-context LLM inference faces a fundamental conflict: head-adaptive compression algorithms (e.g., Top-$p$ nucleus sampling) offer superior accuracy by dynamically fluctuating memory budgets, yet modern inference engines (e.g., vLLM) demand rigid, static memory patterns to leverage CUDA Graphs and PagedAttention. We resolve this ``Static-Dynamic'' mismatch with HARD-KV, a unified framework that that bridges dynamic selection with rigid system constraints. HARD-KV introduces a Cascade Cache hierarchy, managing the token lifecycle across dense, sparse, and condensed tiers. Crucially, we propose a Logits Calibration mechanism that normalizes diverse importance metrics into a unified probability space, enabling consistent Top-$p$ budgeting across heterogeneous heads. To bridge the efficiency gap, we offer a system-level solution, which rewrites fragmented, dynamic indices into contiguous physical layouts compatible with high-performance inference engine. Extensive experiments on math-reasoning benchmarks (AIME, U-Math) verify that HARD-KV achieves up to 2$\times$ throughput improvement over static baselines while maintaining high-fidelity generation in 10k+ token scenarios. Code is available at https://github.com/SuDIS-ZJU/HARDInfer.

cs.LG↗
arXiv2026

Model Selection with Regression and Representational Similarity Analysis for Linear and Nonlinear Data

In cognitive psychology and neuroscience, adjudicating between competing theoretical models is a common methodological challenge. Researchers often rely on either first-order direct mapping approaches (e.g., linear regression) or second-order abstraction methods (e.g., Representational Similarity Analysis [RSA]). However, it remains unclear whether or how the nature of the underlying data and feature characteristics affect the performance of these methods. Here, we systematically evaluated regression, RSA, and Pattern Component Modeling (PCM) across distinct data-generating schemes, including first-order linear mappings and geometry-to-first-order transformations with either linear or nonlinear sigmoid readouts, using both univariate behavioral and multivariate fMRI spatial-pattern simulations. Our results suggest that the relative performance of these methods depends on the underlying generative mechanism. Under linear generative assumptions, regression and PCM showed higher model-selection accuracy than RSA. Under nonlinear but order-preserving transformations, rank-based RSA showed an advantage over regression and PCM. We also found that feature multicollinearity affected these methods differently across generative schemes, and that orthogonalizing the predictor space via principal component analysis (PCA) reduced several collinearity-related differences. Finally, analyses of empirical datasets were consistent with the simulation results under approximately linear conditions, with regression showing clearer model discrimination than RSA. Overall, these findings suggest that the relative performance of regression, RSA, and PCM depends on the form of the mapping between features and responses, as well as on the structure of the feature space.

stat.ME↗
arXiv2026

OPRD: On-Policy Representation Distillation

On-policy distillation (OPD) supervises the student exclusively in the output space by matching next-token distributions. This paradigm suffers from two limitations: (i) a high-variance gradient estimator whose signal-to-noise ratio collapses as the student approaches the teacher, and (ii) an LM-head information bottleneck that discards the teacher's intermediate hidden states. We propose On-Policy Representation Distillation (OPRD), the first method to lift on-policy distillation into the hidden-state space. OPRD aligns student and teacher representations across selected layers on the same on-policy rollouts, providing dense, deterministic, per-layer supervision while bypassing the LM head entirely. Theoretically, OPRD provides a deterministic per-sample gradient, removing the token-level estimation variance that plagues OPD, and exposes structural information that any output-space objective necessarily discards. Empirically, OPRD closes the student-teacher gap on competition mathematics benchmarks (AIME 2024, AIME 2025, and AIMO), where every output-space baseline plateaus below the teacher, while training 1.44x faster and using up to 54% less memory. We further extend OPRD to the cross-architecture setting via OPRD-Bridge. By exploiting the observation that heterogeneous models share a low-rank representational structure, we construct a frozen projector pair that aligns representations across arbitrary depth and width mismatches, shifting the alignment from the output space (which depends on a shared vocabulary) to the representation space. We validate OPRD-Bridge on both cross-architecture (Qwen3-4B -> Qwen3-1.7B-Base) and cross-tokenizer (Phi-4-mini-reasoning -> Qwen3-1.7B-Base) settings, demonstrating successful knowledge transfer even when the vocabulary-based alignment channel is unavailable. Code: https://github.com/ShenzhiYang2000/OPRD.

cs.LG↗
arXiv2026

A DeepLearning Framework for Dynamic Estimation of Origin-Destination Sequence

OD matrix estimation is a critical problem in the transportation domain. The principle method uses the traffic sensor measured information such as traffic counts to estimate the traffic demand represented by the OD matrix. The problem is divided into two categories: static OD matrix estimation and dynamic OD matrices sequence(OD sequence for short) estimation. The above two face the underdetermination problem caused by abundant estimated parameters and insufficient constraint information. In addition, OD sequence estimation also faces the lag challenge: due to different traffic conditions such as congestion, identical vehicle will appear on different road sections during the same observation period, resulting in identical OD demands correspond to different trips. To this end, this paper proposes an integrated method, which uses deep learning methods to infer the structure of OD sequence and uses structural constraints to guide traditional numerical optimization. Our experiments show that the neural network(NN) can effectively infer the structure of the OD sequence and provide practical constraints for numerical optimization to obtain better results. Moreover, the experiments show that provided structural information contains not only constraints on the spatial structure of OD matrices but also provides constraints on the temporal structure of OD sequence, which solve the effect of the lagging problem well.

cs.LG↗
arXiv2026

Large-Scale OD Matrix Estimation with A Deep Learning Method

The estimation of origin-destination (OD) matrices is a crucial aspect of Intelligent Transport Systems (ITS). It involves adjusting an initial OD matrix by regressing the current observations like traffic counts of road sections (e.g., using least squares). However, the OD estimation problem lacks sufficient constraints and is mathematically underdetermined. To alleviate this problem, some researchers incorporate a prior OD matrix as a target in the regression to provide more structural constraints. However, this approach is highly dependent on the existing prior matrix, which may be outdated. Others add structural constraints through sensor data, such as vehicle trajectory and speed, which can reflect more current structural constraints in real-time. Our proposed method integrates deep learning and numerical optimization algorithms to infer matrix structure and guide numerical optimization. This approach combines the advantages of both deep learning and numerical optimization algorithms. The neural network(NN) learns to infer structural constraints from probe traffic flows, eliminating dependence on prior information and providing real-time performance. Additionally, due to the generalization capability of NN, this method is economical in engineering. We conducted tests to demonstrate the good generalization performance of our method on a large-scale synthetic dataset. Subsequently, we verified the stability of our method on real traffic data. Our experiments provided confirmation of the benefits of combining NN and numerical optimization.

cs.AI↗
arXiv2026

Sharp and unified $L^2$ error estimates for the nonsymmetric Nitsche method on convex polytopes

Nitsche's method weakly imposes Dirichlet boundary conditions, but its nonsymmetric variant has long shown a gap between theory and computation: the classical $L^2$~analysis under $H^{k+1}$~regularity predicts a half-order convergence loss, whereas numerical experiments on smooth test problems consistently produce the optimal rate. Whether this discrepancy reflects a limitation of the analysis or an essential feature of the method has remained an open question. On bounded convex polytopes in two and three dimensions, we prove a unified, regularity-dependent $L^2$~error estimate valid across the entire penalty scale $h^{-α}$: \begin{align*} \|u-u_h\|_{L^2(Ω)}\le C h^r |u|_{W^{k+1,p}(Ω)},\qquad r=\min\bigl\{k+1,\,k+\max\{1,α\}-1/p\bigr\}. \end{align*} Numerical experiments in two and three dimensions, on a one-parameter family of manufactured solutions with tunable regularity, demonstrate the sharpness of the estimate and resolve the open question. First, under merely $H^{k+1}$~regularity the half-order loss is essential; second, the optimal convergence consistently observed on smooth test problems is therefore explained by their full $W^{k+1,\infty}$~regularity, not by a limitation of the standard analysis. The theory identifies $α\ge 1+1/p$ or $p=\infty$ as the sharp threshold for recovering the optimal rate $h^{k+1}$, and the experiments confirm this if-and-only-if condition in both dimensions.

math.NA↗
arXiv2026

Smart Picks in the Dark: Towards Efficient RLVR for Reasoning via Tracing Metacognitive Pivots

Reinforcement learning with verifiable rewards (RLVR) has greatly advanced large reasoning models (LRMs), but it requires timely training on a huge fully-annotated dataset. To this end, data-efficient RLVR methods have been widely studied from two perspectives: (i) data selection methods identify a small subset of "golden" samples that yield near-full-data performance, but they rely on a pre-existing pool of labeled data. (ii) unsupervised RLVR methods train the model using its own internal supervision signals on large-scale unlabeled data, yet they exhibit suboptimal performance. Accordingly, we investigate the "pick in the dark" setup for RLVR, which aims to select, without prior supervision, unlabeled samples that are most beneficial for training and worthy of annotation. Through systematic analysis, we demonstrate that smart picks hinge on a well-calibrated uncertainty estimator to enable strategic partitioning of data for adaptive training regimes. Building on this insight, we propose PivotTrace, a three-way data triage framework that leverages attention dynamics to trace metacognitive pivots during reasoning. By precisely quantifying uncertainty through pivot density, PivotTrace achieves automated data routing to synergistically maximize both annotation and training efficiency. Empirically, PivotTrace surpasses the fully supervised LRM with only 29.3% annotated samples and 2.75 faster convergence.

cs.LG↗
arXiv2026

GeoMin: Data-Efficient Semi-Supervised RLVR via Geometric Distribution Modeling

Reinforcement learning with verifiable rewards (RLVR) significantly advances LLM reasoning, yet it faces a dilemma: standard supervised scaling is throttled by high annotation costs, while unsupervised alternatives suffer from severe model collapse. Recent semi-supervised RLVR methods address this by using a small labeled set to guide unlabeled data, achieving a promising trade-off between training efficacy and annotation cost. However, they suffer from a severe data-efficiency bottleneck due to the reliance on coarse performance heuristics, leaving a vast majority of valuable instances underutilized. To this end, we propose GeoMin, which models global feature distributions on labeled data to decode the structural discrepancy between correct and incorrect rollouts, thereby establishing a robust prior to assess the reliability of self-reward signals and fully unleash the potential of unlabeled data. Empirically, GeoMin outperforms the strongest baselines by +4.1% and even surpasses fully supervised models with only 10% of the annotations, demonstrating remarkable data efficiency.

cs.LG↗
arXiv2026

Deft Scheduling of Dynamic Cloud Workflows with Varying Deadlines via Mixture-of-Experts

Workflow scheduling in cloud computing demands the intelligent allocation of dynamically arriving, graph-structured workflows with varying deadlines onto ever-changing virtual machine resources. However, existing deep reinforcement learning (DRL) schedulers remain limited by rigid, single-path inference architectures that struggle to handle diverse scheduling scenarios. We introduce $\textbf{DEFT}$ ($\textbf{D}$eadline-p$\textbf{E}$rceptive Mixture-o$\textbf{F}$-Exper$\textbf{t}$s), an innovative DRL policy architecture that leverages a specialized mixture of experts, each trained to manage different levels of deadline tightness. To our knowledge, DEFT is the first to introduce and validate a Mixture-of-Experts architecture for dynamic cloud workflow scheduling. By adaptively routing decisions through the most appropriate experts, DEFT is capable of meeting a broad spectrum of deadline requirements that no single expert can achieve. Central to DEFT is a $\textbf{graph-adaptive}$ gating mechanism that encodes workflow DAGs, task states, and VM conditions, using cross-attention to guide expert activation in a fine-grained, deadline-sensitive manner. Experiments on dynamic cloud workflow benchmarks demonstrate that DEFT significantly reduces execution cost and deadline violations, outperforming multiple state-of-the-art DRL baselines.

cs.AI↗
arXiv2026

Set-Supervised Diffusion Policy: Learning Action-Chunking Diffusion through Corrections

Diffusion policies have recently emerged as a powerful framework for robotic manipulation. However, like other behavior cloning methods, they remain vulnerable to distributional shift, often requiring human-in-the-loop interventions to correct failures during deployment. These interactions naturally provide paired supervision in the form of the robot's undesired actions and the human teacher's corrective actions. Yet existing data aggregation pipelines and standard behavior cloning losses largely ignore this negative signal from undesired actions, leading to overfitting to teacher's actions and an increasing reliance on costly expert data. To address this limitation, we propose Set-Supervised Diffusion Policy (SDP), a novel learning framework that utilizes contrastive action-chunk data to train diffusion policies from human corrections. From paired positive and negative action-chunks, SDP constructs a set of desired action-chunks and designs a training pipeline that encourages the diffusion policy to align with the set. Through extensive experiments across multiple robotic manipulation tasks, we demonstrate that SDP consistently improves policy performance, with particularly strong gains in robustness to noisy data. Moreover, SDP induces high-quality aggregated datasets, enabling more efficient and reliable policy learning from human-in-the-loop corrections. Our code is available at https://set-supervised-diffusion-policy.github.io/.

cs.RO↗