arXiv subjects

Zijun Chen

Publications and source records attributed to Zijun Chen.

At least 19 recordsLinked to original sources

arXiv2026

StrategyBench: Evaluating Explicit Strategy Induction in Large Language Models

As large language models are increasingly used in data-scarce and evolving task scenarios, few-shot in-context learning (ICL) has become a key paradigm for task adaptation. However, direct ICL often uses a small set of examples without explicitly abstracting task rules, making it sensitive to example construction. In contrast, human learners often reduce such sensitivity by first summarizing task rules from examples and then applying them to new instances. To evaluate this ability, we propose StrategyBench, which selects strategy-inducible tasks from BIG-Bench, constructs reference strategies, and defines evaluation metrics along two dimensions: strategy quality and downstream utility. We further analyze strategy induction from three perspectives: task variation, model configuration, and adaptation setting, covering category-wise differences, generator-executor choices, demonstration design, and SFT-based adaptation. Experiments show that explicit strategy utility differs substantially across task categories and depends on both strategy generation and execution conditions. The benchmark is released at: https://anonymous.4open.science/r/StrategyBench-D53C.

cs.AI↗

arXiv2026

Precision quantum simulation of magnon spectra and interactions

Quantum simulation promises to advance materials discovery by accurately simulating complex states of matter, their microscopic excitations, and macroscopic response functions. The central challenge in resolving the underlying interacting dynamics is to combine high-fidelity evolution with the sophisticated control necessary to manipulate individual quasi-particles in quantum many-body states. Here, we report on high-precision simulation of both linear and non-linear response functions in a 2D XY spin-1/2 magnet using an analog-digital superconducting processor of up to 97 qubits. By interleaving digital gates with analog evolution precisely characterized via Hamiltonian learning, we selectively excite magnons at tunable energy densities. Measuring first the linear magnon response -- a central probe in neutron-scattering experiments -- we extract temperature-dependent spectra and lifetimes. Our results reveal stark variations in magnon decay rates across the Brillouin zone, with enhancement near van Hove singularities and suppression for edge-localized modes. Next, we perform a suite of nonlinear measurements, including the study of self-scattering mechanisms, as well as pump-probe spectroscopy to directly characterize the magnon interactions. While matrix-product state simulations capture the dynamics well in either small systems or at low temperatures, their predictions become inaccurate away from these limits. This work demonstrates precise simulation of the interacting dynamics in quantum magnets, and provides key insights into quasi-particles and their microscopic scattering mechanisms.

quant-ph↗

arXiv2026

Tighter Regret Bounds for Contextual Action-Set Reinforcement Learning

We study episodic reinforcement learning with fixed reward and transition functions, but with episode-dependent admissible action sets that are observed at the start of each episode. Performance is measured by cumulative regret against the episode-wise optimal value, $\sum_{k=1}^K [V^{*,M^k} - V^{\pi^k,M^k}]$, where $M^k$ represents the action context in the $k$-th episode. We show that the MVP algorithm naturally extends to this framework and enjoys strong theoretical guarantees. In particular, we establish a minimax regret bound of $\widetilde{O}(\sqrt{SAH^3K\log L})$ for adversarial contexts, where $L$ denotes the number of possible contexts. This result implies a regret bound of $\widetilde{O}(\sqrt{SAH^3K})$ for stochastic contexts. We further translate the stochastic regret guarantee into a sample complexity bound of $\widetilde{O}(SAH^3/\epsilon^2)$ for a fixed context distribution. In addition, we derive a gap-dependent regret bound of \[ \widetilde O\left( \inf_{p\in [0,1)} \left( \frac{1}{\Delta_{\min}^{p}} + pK\Delta_{\min}^{p} \right)\log K \cdot \mathrm{poly}(S,A,H) \right), \] where $\Delta_{\min}^{p}$ is the global $p$-trimmed positive-gap floor over suboptimal $(h,s,a)$ triples. This bound can substantially improve upon the minimax rate when the relevant suboptimality gaps are large.

cs.LG↗

arXiv2026

A characterization of 4-connected graphs with no 6-wheel minor

For each integer $n\geq 3$, let $W_n$ denote the wheel graph obtained by connecting a single vertex to all vertices of a cycle of length $n$. In particular, $W_6$ is obtained from the Petersen graph by contracting three edges incident with a common vertex. In this paper, we determine all $4$-connected graphs that do not contain $W_6$ as a minor.

math.CO↗

arXiv2026

An excluded minor theorem for the 6-wheel

For each integer $n \geq 3$, the wheel graph $W_n$ is defined as the graph obtained by connecting a single vertex to all vertices of a cycle of length $n$. In particular, $W_6$ can be uniquely obtained from the Petersen graph by contracting three edges incident to a common vertex. Gubser provided a characterization of all 3-connected planar $W_6$-minor-free graphs. In this paper, we complete the characterization of $W_6$-minor-free graphs by determining the 3-connected nonplanar cases.

math.CO↗

arXiv2026

LongAct: Harnessing Intrinsic Activation Patterns for Long-Context Reinforcement Learning

Reinforcement Learning (RL) has emerged as a critical driver for enhancing the reasoning capabilities of Large Language Models (LLMs). While recent advancements have focused on reward engineering or data synthesis, few studies exploit the model's intrinsic representation characteristics to guide the training process. In this paper, we first observe the presence of high-magnitude activations within the query and key vectors when processing long contexts. Drawing inspiration from model quantization -- which establishes the criticality of such high-magnitude activations -- and the insight that long-context reasoning inherently exhibits a sparse structure, we hypothesize that these weights serve as the pivotal drivers for effective model optimization. Based on this insight, we propose LongAct, a strategy that shifts from uniform to saliency-guided sparse updates. By selectively updating only the weights associated with these significant activations, LongAct achieves an approximate 8% improvement on LongBench v2 and enhances generalization on the RULER benchmark. Furthermore, our method exhibits remarkable universality, consistently boosting performance across diverse RL algorithms such as GRPO and DAPO. Extensive ablation studies suggest that focusing on these salient features is key to unlocking long-context potential.

cs.LG↗

arXiv2026

Curriculum-Based Soft Actor-Critic for Multi-Section R2R Tension Control

Precise tension control in roll-to-roll (R2R) manufacturing is difficult under varying operating conditions and process uncertainty. This paper presents a curriculum-based Soft Actor-Critic (SAC) controller for multi-section R2R tension control. The policy is trained in three phases with progressively wider reference ranges, from 27 to 33 N to the full operating envelope of 20 to 40 N, so it can generalize across nominal and disturbed conditions. On a three-section R2R benchmark, the learned controller achieves accurate tracking in nominal operation and handles large disturbances, including 20 N to 40 N step changes, with a single policy and no scenario-specific retuning. These results indicate that curriculum-trained SAC is a practical alternative to model-based control when system parameters vary and process uncertainty is significant.

eess.SY↗

arXiv2026

LongR: Unleashing Long-Context Reasoning via Reinforcement Learning with Dense Utility Rewards

Reinforcement Learning has emerged as a key driver for LLM reasoning. This capability is equally pivotal in long-context scenarios--such as long-dialogue understanding and structured data analysis, where the challenge extends beyond consuming tokens to performing rigorous deduction. While existing efforts focus on data synthesis or architectural changes, recent work points out that relying solely on sparse, outcome-only rewards yields limited gains, as such coarse signals are often insufficient to effectively guide the complex long-context reasoning. To address this, we propose LongR, a unified framework that enhances long-context performance by integrating a dynamic "Think-and-Read" mechanism, which interleaves reasoning with document consultation, with a contextual density reward based on relative information gain to quantify the utility of the relevant documents. Empirically, LongR achieves a 9% gain on LongBench v2 and consistent improvements on RULER and InfiniteBench, demonstrating robust efficiency in navigating extensive contexts. Furthermore, LongR consistently enhances performance across diverse RL algorithms (e.g., DAPO, GSPO). Finally, we conduct in-depth analyses to investigate the impact of reasoning chain length on efficiency and the model's robustness against distractors.

cs.CL↗

arXiv2026

Achieving $\varepsilon^{-2}$ Dependence for Average-Reward Q-Learning with a New Contraction Principle

We present the convergence rates of synchronous and asynchronous Q-learning for average-reward Markov decision processes, where the absence of contraction poses a fundamental challenge. Existing non-asymptotic results overcome this challenge by either imposing strong assumptions to enforce seminorm contraction or relying on discounted or episodic Markov decision processes as successive approximations, which either require unknown parameters or result in suboptimal sample complexity. In this work, under a reachability assumption, we establish optimal $\widetilde{O}(\varepsilon^{-2})$ sample complexity guarantees (up to logarithmic factors) for a simple variant of synchronous and asynchronous Q-learning that samples from the lazified dynamics, where the system remains in the current state with some fixed probability. At the core of our analysis is the construction of an instance-dependent seminorm and showing that, after a lazy transformation of the Markov decision process, the Bellman operator becomes one-step contractive under this seminorm.

cs.LG↗

arXiv2026

Hilbert space signatures of non-ergodic glassy dynamics

Disorder in quantum many-body systems can drive transitions between ergodic and non-ergodic phases, yet the nature--and even the existence--of these transitions remains intensely debated. Using a two-dimensional array of superconducting qubits, we study an interacting spin model at finite temperature in a disordered landscape, tracking dynamics both in real space and in Hilbert space. Over a broad disorder range, we observe an intermediate non-ergodic regime with glass-like characteristics: physical observables become broadly distributed and some, but not all, degrees of freedom are effectively frozen. The Hilbert-space return probability shows slow power-law decay, consistent with finite-temperature quantum glassiness. In the same regime, we detect the onset of a finite Edwards-Anderson order parameter and the disappearance of spin diffusion. By contrast, at lower disorder, spin transport persists with a nonzero diffusion coefficient. Our results show that there is a transition out of the ergodic phase in two-dimensional systems.

quant-ph↗

arXiv2025

A Graph Neural Network with Auxiliary Task Learning for Missing PMU Data Reconstruction

In wide-area measurement systems (WAMS), phasor measurement unit (PMU) measurement is prone to data missingness due to hardware failures, communication delays, and cyber-attacks. Existing data-driven methods are limited by inadaptability to concept drift in power systems, poor robustness under high missing rates, and reliance on the unrealistic assumption of full system observability. Thus, this paper proposes an auxiliary task learning (ATL) method for reconstructing missing PMU data. First, a K-hop graph neural network (GNN) is proposed to enable direct learning on the subgraph consisting of PMU nodes, overcoming the limitation of the incompletely observable system. Then, an auxiliary learning framework consisting of two complementary graph networks is designed for accurate reconstruction: a spatial-temporal GNN extracts spatial-temporal dependencies from PMU data to reconstruct missing values, and another auxiliary GNN utilizes the low-rank property of PMU data to achieve unsupervised online learning. In this way, the low-rank properties of the PMU data are dynamically leveraged across the architecture to ensure robustness and self-adaptation. Numerical results demonstrate the superior offline and online performance of the proposed method under high missing rates and incomplete observability.

eess.SY↗

arXiv2025

GeoBench: Rethinking Multimodal Geometric Problem-Solving via Hierarchical Evaluation

Geometric problem solving constitutes a critical branch of mathematical reasoning, requiring precise analysis of shapes and spatial relationships. Current evaluations of geometric reasoning in vision-language models (VLMs) face limitations, including the risk of test data contamination from textbook-based benchmarks, overemphasis on final answers over reasoning processes, and insufficient diagnostic granularity. To address these issues, we present GeoBench, a hierarchical benchmark featuring four reasoning levels in geometric problem-solving: Visual Perception, Goal-Oriented Planning, Rigorous Theorem Application, and Self-Reflective Backtracking. Through six formally verified tasks generated via TrustGeoGen, we systematically assess capabilities ranging from attribute extraction to logical error correction. Experiments reveal that while reasoning models like OpenAI-o3 outperform general MLLMs, performance declines significantly with increasing task complexity. Key findings demonstrate that sub-goal decomposition and irrelevant premise filtering critically influence final problem-solving accuracy, whereas Chain-of-Thought prompting unexpectedly degrades performance in some tasks. These findings establish GeoBench as a comprehensive benchmark while offering actionable guidelines for developing geometric problem-solving systems.

cs.CV↗

arXiv2025

Observation of disorder-induced superfluidity

The emergence of states with long-range correlations in a disordered landscape is rare, as disorder typically suppresses the particle mobility required for long-range coherence. But when more than two energy levels are available per site, disorder can induce resonances that locally enhance mobility. Here we explore phases arising from the interplay between disorder, kinetic energy, and interactions on a superconducting processor with qutrit readout and control. Compressibility measurements distinguish an incompressible Mott insulator from surrounding compressible phases and reveal signatures of glassiness, reflected in non-ergodic behavior. Spatially-resolved two-point correlator measurements identify regions of the phase diagram with a non-vanishing condensate fraction. We also visualize the spectrum by measuring the dynamical structure factor. A linearly-dispersing phonon mode materializes in the superfluid, appearing even when disorder is introduced to the clean Mott insulator. Our results provide strong experimental evidence for disorder-induced superfluidity.

quant-ph↗

arXiv2025

A semicircle law for the normalized Laplacian of sparse random graphs

We study the limiting spectral distribution of the normalized Laplacian $\mathcal L$ of an Erd\H{o}s-R\'enyi graph $G(n,p)$. To account for the presence of isolated vertices in the sparse regime, we define $\mathcal L$ using the Moore-Penrose pseudoinverse of the degree matrix. Under this convention, we show that the empirical spectral distribution of a suitably normalized $\mathcal L$ converges weakly in probability to the semicircle law whenever $np\to\infty$, thereby providing a rigorous justification of a prediction made in (Akara-pipattana and Evnin, 2023). Moreover, if $np>\log n+\omega(1)$, so that $G(n,p)$ has no isolated vertices with high probability, the same conclusion holds for the standard definition of $\mathcal L$. We further strengthen this result to almost sure convergence when $np=\Omega(\log n)$. Finally, we extend our approach to the Chung-Lu random graph model, where we establish a semicircle law for $\mathcal L$ itself, improving upon (Chung, Lu, and Vu 2003), which obtained the semicircle law only for a proxy matrix.

math.PR↗

arXiv2025

Magic state cultivation on a superconducting quantum processor

Fault-tolerant quantum computing requires a universal gate set, but the necessary non-Clifford gates represent a significant resource cost for most quantum error correction architectures. Magic state cultivation offers an efficient alternative to resource-intensive distillation protocols; however, testing the proposal's assumptions represents a challenging departure from quantum memory experiments. We present an experimental study of magic state cultivation on a superconducting quantum processor. We implement cultivation, including code-switching into a surface code, and develop a fault-tolerant measurement protocol to bound the magic state fidelity. Cultivation reduces the error by a factor of 40, with a state fidelity of 0.9999(1) (retaining 8% of attempts). Our results experimentally establish magic state cultivation as a viable solution to one of quantum computing's most significant challenges.

quant-ph↗

arXiv2025

Quantum-Classical Separation in Bounded-Resource Tasks Arising from Measurement Contextuality

The prevailing view is that quantum phenomena can be harnessed to tackle certain problems beyond the reach of classical approaches. Quantifying this capability as a quantum-classical separation and demonstrating it on current quantum processors has remained elusive. Using a superconducting qubit processor, we show that quantum contextuality enables certain tasks to be performed with success probabilities beyond classical limits. With a few qubits, we illustrate quantum contextuality with the magic square game, as well as quantify it through a Kochen--Specker--Bell inequality violation. To examine many-body contextuality, we implement the N-player GHZ game and separately solve a 2D hidden linear function problem, exceeding classical success rate in both. Our work proposes novel ways to benchmark quantum processors using contextuality-based algorithms.

quant-ph↗

arXiv2025

An LLM-Assisted Multi-Agent Control Framework for Roll-to-Roll Manufacturing Systems

Roll-to-roll manufacturing requires precise tension and velocity control to ensure product quality, yet controller commissioning and adaptation remain time-intensive processes dependent on expert knowledge. This paper presents an LLM-assisted multi-agent framework that automates control system design and adaptation for R2R systems while maintaining safety. The framework operates through five phases: system identification from operational data, automated controller selection and tuning, sim-to-real adaptation with safety verification, continuous monitoring with diagnostic capabilities, and periodic model refinement. Experimental validation on a R2R system demonstrates successful tension regulation and velocity tracking under significant model uncertainty, with the framework achieving performance convergence through iterative adaptation. The approach reduces manual tuning effort while providing transparent diagnostic information for maintenance planning, offering a practical pathway for integrating AI-assisted automation in manufacturing control systems.

eess.SY↗

arXiv2025

Panel-Scale Reconfigurable Photonic Interconnects for Scalable AI Computation

Panel-scale reconfigurable photonic interconnects on a glass substrate up to 500-mm x 500-mm or larger are envisioned by proposing a novel photonic switch fabric that enables all directional panel-edge-to-panel-edge reach without the need for active repeaters while offering high communication bandwidth, planar-direction reconfigurability, low energy consumption, and compelling data bandwidth density for heterogeneous integration of an in-package AI computing system on a single glass-substrate photonic interposer exceeding thousands of centimeters square. The proposed approach focuses on reconfigurable photonic interconnects, which are integration-compatible with commercial processor chiplets and 3D high-bandwidth memory (HBM) stacks on a large-area glass substrate, to create a novel panel-scale heterogeneously integrated interposer or package enabling low-energy and high-capacity wavelength-division-multiplexing (WDM) optical data links using advanced high-speed optical modulators, broadband photodetectors, novel optical crossbar switches with multi-layer waveguides, and in-package frequency comb sources.

eess.SY↗